Aquaculture nutrition, tập 16, số 4, 2010

Aquaculture Nutrition
doi: 10.1111/j.1365-2095.2009.00669.x

2010 16; 335–342

..........................................................................................

1
1

1,2

1

1

1

State Key Laboratory of Freshwater Ecology and Biotechnology, Institute of Hydrobiology, the Chinese Academy of Sciences,
Wuhan, Hubei, China; 2 Aquaculture Divisions, E-Institute of Shanghai Universities, Shanghai, China

The effects of aflatoxin B1 (AFB1) on growth, physiological
responses and histological changes were investigated in juvenile gibel carp (Carassius auratus gibelio). Triplicate groups of
gibel carp (3.53 ± 0.02 g) were fed seven semipurified diets
(Diet 1 to 7) containing 3.20, 5.37, 7.08, 9.55, 12.70, 17.90 and
28.60 lg AFB1 kg)1 diet for 3 months. The results showed fish
weight gain fed Diet 6 was 112.6% of that of control group
(Diet 1) after 3 months, but there was no significant difference
of weight gain between fish fed Diet 7 and the control group.
Alanine aminotransferase (ALT) of fish hepatopancreas fed
Diet 7 was significantly higher than the control group

(P < 0.05), but no significant difference was observed in ALT
activities of the fish fed with more than 10 lg AFB1 kg)1 (Diet
4, 5, 6 and 7). No significant histological lesions were identified
between the control and increasing AFB1 treatments. AFB1
accumulated in hepatopancreas was logarithmically related to
the dietary AFB1 levels, and AFB1 also accumulated in muscles and ovaries of gibel carp fed Diet 3 to Diet 7. The present
results indicated that fish fed with more than 10 lg AFB1 kg)1
diet showed impaired physiological responses and more AFB1
residue of muscles and ovaries above the safety limitation of
European Union.
KEY WORDS: aflatoxin B1, alanine aminotransferase, Carassius
auratus gibelio, residue, weight gain

Received 18 December 2008, accepted 13 March 2009
Correspondence: Dr Dong Han, State Key Laboratory of Freshwater and
Biotechnology, Institute of Hydrobiology, the Chinese Academy of
Sciences, Wuhan, Hubei, 430072, China. E-mail:

Aflatoxin, a polycyclic aromatic hydrocarbon, is produced
mainly by Aspergillus flavus and Aspergillus parasiticus,

..............................................................................................

Ó 2009 The Authors
Journal compilation Ó 2009 Blackwell Publishing Ltd

which occurs naturally in several important plant feedstuffs.
Aflatoxin B1 (AFB1) is the most toxic compound produced
by these moulds. The toxicity of AFB1 has been best reported
for poor growth, hepatotoxic, nephrotoxic, carcinogenic,

mutagenic, teratogenic and cytotoxic properties (Halver
1969; Jantrarotai & Lovell 1990; Cha´vez-Sa´nchez et al. 1994;
Sahoo et al. 1996; Santacroce et al. 2008). In studies of carcinogenic effect, the liver of fish is one of the organs most
affected by dietary AFB1 (Hendricks et al. 1985; McKean
et al. 2006). AFB1 has been found to be a potent immunomodulator in endotherms (Ottinger & Kaattari 1998; Sahoo
& Mukherjee 2001).
AFB1 is well known for the highest potency as a toxin
and is classified as group I carcinogen by International
Agency for Research on Cancer. The consumption of AFB1
contaminated feed by fish brings contamination of products, which may result in secondary contamination for
human from eating meat with residues of AFB1
(Madhusudhanan et al. 2006). Globally, there are strict
regulations with regard to the safe limits of AFB1 in domestic
animal feed and food items (Gowda et al. 2007). The
European Union established 2 lg kg)1 as the maximum
allowable concentration of AFB1 in human food (Food and
Agriculture Organization (FAO) 2004) and 5 lg kg)1 as the
limit of AFB1 in feed (van Eijkeren et al. 2006). Under FDA
regulations, upper limit of most feed ingredients and nonruminant feeds is 20 lg aflatoxin kg)1 (Food and Drug
Administration (FDA) 1989). In many countries including
China, there are no official safe levels for aflatoxins in fish
feeds. Although AFB1 has been studied in different fishes,
such as salmonoids (Hendricks et al. 1977), channel catfish
(Jantrarotai & Lovell 1990) and Nile tilapia (Tuan et al.
2002), most of these reports were focused on the high AFB1
levels. Actually, dietary AFB1 levels could be very low due to
good care of dietary ingredients. Therefore, it is very
important and necessary to make clear the effect of low
dietary AFB1 levels on fish.

Gibel carp, an omnivorous species, is an important aquaculture species in China and the annual production is more
than 3 million tonnes. Increased use of plant ingredients in
aquafeed for gibel carp has intensified the potential for
aflatoxicosis. The purpose of the present study is to investigate the effects of low dietary AFB1 levels on growth, physiological, histological characteristics and tissue AFB1 residue
in gibel carp.

Table 2 Supplemented and determined aflatoxin B1 levels of the
experimental diets
Experimental
diets

Diet
1

Diet
2

Diet
3

Diet
4

Diet
5

Diet
6

Diet
7

Supplemented
AFB1 (lg kg)1)
Determined
AFB1 (lg kg)1)

0

1

5

10

20

40

80

3.20

5.37

7.08

12.70

17.90

28.60

9.55

fish oil were used as the main protein and lipid source. Diets
were made into pellets (2 mm, diameter), oven-dried at 60 °C
and stored at 4 °C until fed.
Purified aflatoxin B1 was purchased from Sigma (St. Louis,
MO, USA). According to the method of Sahoo & Mukherjee
(2003), AFB1 was first dissolved in chloroform and fish oil
was then added into the mixture. After adding oil, the
chloroform was allowed to evaporate.
Seven experimental diets (Diet 1-7) were formulated to
contain 3.20, 5.37, 7.08, 9.55, 12.70, 17.90 and 28.60 lg
AFB1 kg)1 diet, which were isonitrogenous (394.7 g kg)1
crude protein), isoenergetic (16.75 kJ g)1 diet) and isolipidic
(85.4 g kg)1 crude lipid). Diet 1 is the control diet without
supplementing purified aflatoxin B1. The formulation of the
basal diet and the determined AFB1 levels of the experimental diets are shown in Tables 1 & 2. White fish meal and
Table 1 Formulation and chemical composition of the basal diet (in
dry weight)

Ingredient
White fishmeal (from USA)
Corn starch
a-Starch
Fish oil
Mineral premix1

Vitamin premix2
Vitamin C
Cellulose
Carboxymethyl cellulose
Chemical composition (in dry matter)
Crude protein
Crude fat
Gross energy (kJ g)1)

Content
(g kg)1)
570.0
220.0
30.0
25.0
50.0
3.9
1.1
70.0
30.0
394.7
85.4
16.75

Mineral premix (mg kg)1 diet): NaCl, 500; MgSO4Æ7H2O, 7500;
NaH2PO4Æ2H2O, 12500; KH2PO4, 16000; Ca (H2PO4) 2ÆH2O, 10000;
FeSO4, 1250; C6H10CaO6Æ5H2O, 1750; ZnSO4Æ7H2O, 176.5; MnSO4Æ
4H2O, 81; CuSO4Æ5H2O, 15.5; CoSO4Æ6H2O, 0.5; KI, 1.5; starch, 225.
2
Vitamin premix (mg kg)1 diet): thiamin, 20; riboflavin, 20; pyridoxine, 20; cyanocobalamine, 2; folic acid, 5; calcium patotheniate,

50; inositol, 100; niacin, 100; biotin, 5; starch, 3226; Vitamin A, 110;
Vitamin D3, 20; Vitamin E, 100; Vitamin K3, 10; Choline chloride,
1100.
1

Gibel carp was obtained from the hatchery of the Institute of
Hydrobiology, the Chinese Academy of Sciences. Before the
experiment, the juveniles were acclimated to the experimental
condition for 2 weeks. During the acclimation, fish were fed
twice daily with the control diet.
The experiment was carried out in a flow-through system
consisting of 21 polythene tanks (60 · 47 · 50 cm, water
volume: 140 L). Water flowing rate into each tank was
0.2 l min)1. During the experiment, water temperature
was maintained at 24 ± 2 °C. The dissolved oxygen content
was kept above 7.5 mg O2 L)1, pH between 7.0 and 7.6,
ammonia-N less than 0.5 mg L)1, and the photoperiod was
12D : 12L with the light period from 0800 to 2000.

Before the experiment, fish were deprived of feed for 1 day.
Twenty-three fish (3.53 ± 0.02 g ind)1) were bulk weighed
and randomly transferred into each tank. During the experiment, the fish were fed to apparent satiation twice a day (0900,
1500), and daily feed intake was recorded. The faeces were
removed by siphoning twice a day just before each feeding.
Fish in each tank were batch-weighed every month after 1-day
food deprivation to evaluate the toxic effect of dietary AFB1
on growth performance. The trial lasted for 3 months.
At the end of the experiment, 10 fish per tank were anesthetized with MS-222, and the length and weight of fish were
measured. Blood was withdrawn from dorsal vessels using the
syringe, and centrifuged at 4000 g (4 °C) for 15 min to get the

serum. The hepatopancreas of four fish were collected on ice,
weighed and stored in liquid nitrogen for enzyme determination. The muscles, hepatopancreas, and ovaries of six fish were
pooled on ice and used for AFB1 residue analysis. Another
three fish per tank were sampled for histological studies.

..............................................................................................

Ó 2009 The Authors
Journal compilation Ó 2009 Blackwell Publishing Ltd Aquaculture Nutrition 16; 335–342


1

Serum analysis Serum superoxide dismutase (SOD) activity
was determined using the method of McCord & Fridovich
(1969). One unit of SOD activity is described as the amount
of sample required to cause 50% inhibition of the rate of
reduction of cytochrome c by O2). Serum catalase (CAT)
activity was measured using the method of Aebi (1984) in
which the initial rate of hydrogen peroxide decomposition is
determined. One unit of catalase activity was defined as the
amount of enzyme that catalysed the decomposition of
1.0 lmol of H2O2 per second.
Hepatopancreas analysis Fish hepatopancreas were thawed,
rinsed with 0.65% physiological saline, and then homogenized in five volumes v/w of ice-cold 0.65% physiological
saline using a hand held glass homogenizer on ice. The
homogenate was centrifuged at 10 000 g for 10 min at 4 °C.
The supernatant was stored at 4 °C as the crude extract. All
assays were conducted within 24 h after extraction.
The enzyme activities of aspartate aminotransferase (AST)

and alanine aminotransferase (ALT) were measured at an
absorbance of 505 nm according to the method of Hørder &
Rej (1981). The protein content of the supernatant was
determined using Bradford assay with bovine serum albumin
as a standard, and the specific activity of the enzymes was
defined as activity unit per mg protein.

enzyme-link immunosorbent assay (ELISA) method using a
commercial assay kit (Brins-LivePro Biotechnology Corporation, Beijing, China). For this determination, the absorbance at 490 nm was recorded using a microplate
spectrophotometer (PowerWave XS; BioTek Instruments
Inc., Winooski, Vermont, USA). The sensitivity of AFB1
measurement was less than 0.01 lg kg)1.

DuncanÕs multiple range test was used to detect the significance of differences of means between groups followed by
one-way analysis of variance (ANOVA). Effects with a probability of P < 0.05 were considered statistically significant.

The hepatopancreas samples were fixed in BouinÕs fluid for
12 h and dehydrated in a graded ethanol series. The hepatopancreas slices were then embedded in paraffin, thin sectioned to 7 lm, stained with hematoxylin and eosin (H & E)
and observed under light microscopy.

During the experiment, the average weight gain (WG) of fish
was significantly affected by dietary aflatoxin B1 concentrations (Table 3). After 1 month, fish WG increased with the
increasing dietary AFB1, and the fish WG fed with highest
AFB1 level (Diet 7) was significantly higher than the low
dietary AFB1 groups (Diet 1 or 2) (P < 0.05). After
2 months, no significant difference was observed in WG
between the highest AFB1 level group (Diet 7) and the others.
At the end of the experiment, WG of the fish fed Diet 7 was
significant lower than that fed Diet 6 (P < 0.05). Table 3
showed condition factor (CF) in the fish fed Diet 6 was significantly higher than CF in the other groups (P < 0.05).

There was no significant effect on hepatosomatic index (HSI)
and feed conversion ratio (FCR) between all groups
(P > 0.05). There was no mortality observed during the
experiment.

Crude protein, lipid and energy content were analysed for
the experimental diets. Dry matter content was determined
by drying to constant weight at 105 °C. Nitrogen content
was analysed by the Kjeldahl method. Crude lipid was
determined by ether extraction using a Soxtec system
(Soxtec System HT, 1043 Extraction Unit, Tecator, Sweden) and energy by bomb calorimeter (Phillipson microbomb calorimeter; Gentry Instruments Inc., Aiken, SC,
USA).
AFB1 measurement of the experimental diets, muscles,
hepatopancreas and ovaries of fish were carried out by the

The activities of catalase (CAT), superoxide dismutase
(SOD), aspartate aminotransferase (AST) and alanine
aminotransferase (ALT) of gibel carp fed with different
dietary AFB1 for 3 months were shown in Table 4. Serum
SOD activities of fish fed with more than 10 lg AFB1 kg)1 diet
(Diet 4, 5, 6 and 7) were significantly higher than the control
diet (P < 0.05). Hepatopancreas ALT activities of fish fed
Diet 7 were significantly higher than the control (P < 0.05),
while the activities of serum CAT and hepatopancreas AST
showed no significant difference between different groups
(P > 0.05).

..............................................................................................

Ó 2009 The Authors

Journal compilation Ó 2009 Blackwell Publishing Ltd Aquaculture Nutrition 16; 335–342


Table 3 Growth performance of gibel carp fed with different concentrations of AFB1 for 3 months (mean ± SE)*
Average weight gain (%)1

Supplemented
AFB1 (lg kg)1)

Initial weight
(g)

1 month

0
1
5
10
20
40
80

3.52
3.53
3.52
3.54
3.52
3.51
3.54

122.9
127.3
128.6
128.8
130.3
132.5
135.7

±
±
±
±
±
±
±

0.01
0.01
0.01
0.003
0.02
0.01
0.01

±
±
±
±
±
±

±

2 months

3.8a
2.0ab
2.1abc
2.3abc
0.9abc
2.8bc
2.5c

371.5
382.6
389.3
387.8
391.8
401.6
391.8

±
±
±
±
±
±
±

12.0a
3.4ab

4.2ab
2.4ab
5.4ab
3.8b
17.9ab

3 months
739.0
746.2
736.4
802.6
774.4
842.3
774.7

±
±
±
±
±
±
±

35.8a
15.8a
10.0a
11.7ab
15.1a
25.9b
12.3a

HSI (%)2

CF3

3.87
3.55
3.34
4.10
3.73
3.50
3.56

1.45
1.53
1.60
1.63
1.56
1.76
1.59

±
±
±
±
±
±
±

0.27

0.14
0.25
0.20
0.33
0.19
0.27

FCR4
±
±
±
±
±
±
±

0.02a
0.05ab
0.03b
0.03b
0.02ab
0.07c
0.03b

1.26
1.16
1.23
1.14
1.15
1.12

1.16

±
±
±
±
±
±
±

0.10
0.01
0.05
0.01
0.01
0.02
0.004

* Different superscript letters within each row represent significant differences (P < 0.05).
Weight gain = 100 · (Final weight)initial weight)/(initial weight).
2
HSI (Hepatosomatic index) = 100 · (liver weight)/(fish weight).
3
CF (Condition factor) = 100 · final weight (g)/[fork length (cm)]3.
4
FCR (Feed conversion ratio) = total dry feed fed (g)/total wet weight gain (g).
1

Table 4 The activities of catalase (CAT), superoxide dismutase
(SOD), aspartate aminotransferase (AST) and alanine aminotransferase (ALT) of gibel carp fed with different concentrations of AFB1

for 3 months (mean ± SE)*
In serum
Supplemented
CAT
AFB1
(lg kg)1) (U ml)1)

SOD
(U ml)1)

0
1
5
10
20
40
80

211.4
238.9
245.2
278.0
285.9
293.0
305.8

2.6
2.0
2.6
1.9

2.0
2.3
2.6

±
±
±
±
±
±
±

0.34
0.29
0.66
0.31
0.13
0.38
0.29

residue in muscle and ovary was 4.08 and 3.16 lg kg)1
(Fig. 3).

In hepatopancreas

±
±
±
±
±

±
±

6.1a
13.9ab
30.7abc
17.1bc
20.1bc
19.8bc
10.4c

AST
ALT
(IU g)1 pro) (IU g)1 pro)
70.1
65.9
69.3
82.4
79.5
67.0
63.5

±
±
±
±
±
±
±

9.3
10.5
13.4
14.4
13.2
4.6
15.7

153.3
156.2
166.2
205.8
205.1
216.6
244.5

±
±
±
±
±
±
±

19.2a
11.6a
4.9a
32.0ab
21.6ab
6.8ab

36.5b

* Different superscript letters within each row represent significant
differences (P < 0.05).

After 3 months, no significant histological lesions were
identified in hepatopancreatic tissues of gibel carp between
the control and different dietary AFB1 groups. Hepatocytes
and nuclei were uniform in size and shape and prominent
cytoplasm was present in most hepatocytes of the control fish
and AFB1 treated fish (Fig. 1).

1

After 3 months, AFB1 residue in fish hepatopancreas
increased significantly and was logarithmically related to
dietary AFB1 (Fig. 2). However, AFB1 residue in muscle and
ovary of fish was not detected in the control group and Diet
2. Groups of Diet 4, 5, 6 and 7 produced more than 2 lg kg)1
of AFB1 residue in muscles and ovaries, and the highest

The phenomenon of hormesis for toxic or harmful substances, a dose–response characterized by a low dose stimulation and a high dose inhibition, has been widely discussed
(Calabrese & Baldwin 1998, 2003; Calabrese 2005, 2008;
Murado & Va´zquez 2007; Belz et al. 2008). In the present
study, weight gain of fish fed Diet 6 was 112.6% of that of
the control group after 3 months, but there was no significant
difference of weight gain between fish fed Diet 7 and the
control group. This growth hormesis response was similar to
the results in chickens (Diaz et al. 2008). However, decreased
growth was reported in many species fed different dietary

AFB1, which is not agreed to the present study (Boonyaratpalin
et al. 2001; Tuan et al. 2002; Madrigal-Santilla´n et al. 2006;
Han et al. 2008). The reasons for the difference were probably that: (1) many reports studied the toxic effects of high
dietary AFB1 (Jantrarotai & Lovell 1990; Cha´vez-Sa´nchez
et al. 1994; Boonyaratpalin et al. 2001; Tuan et al. 2002); (2)
many research species were sensitive to even very low dietary
AFB1 levels (Lovell 1989; Lim et al. 2001); (3) the homesis
could be usually elicited for the low response of 10–20%
(Johnson & Bruunsgaard 1998). Fish fed Diet 7 showed the
significant growth-suppressing effects of AFB1 over time
during the experiment, which suggested that the hormesis
response could change as experimental time went on. The
condition factor (CF) expressed the condition of fish, such as
the degree of well being, plumpness or fatness, and determined from observed weights and fork length. In this study,

..............................................................................................

Ó 2009 The Authors
Journal compilation Ó 2009 Blackwell Publishing Ltd Aquaculture Nutrition 16; 335–342


1

(a)

50 µm

(b)

(d)

(c)
50 µm

(e)

50 µm

50 µm

50 µm

(f)
50 µm

(g) 50 µm

Figure 1 Hepatopancreas of gibel carp
fed with (a) the control diet, (b) Diet 2,
(c) Diet 3, (d) Diet 4, (e) Diet 5, (f ) Diet
6 and (g) Diet 7 for 3 months. H&E,
Bar = 50 lm.

there was a significant increase of CF in fish fed Diet 3 and
above and fish fed Diet 6 showed the highest CF. It was
suggested that fish fed Diet 3 and above grew shorter in
shape than the control group.
In this study, FCR was not significantly affected by the
dietary AFB1 levels. Similar findings were reported for Nile
tilapia that FCR was not affected by AFB1 levels as high as

30 mg kg)1 (Cha´vez-Sa´nchez et al. 1994). The different
finding in tilapia was reported that FCR of the control was

significantly lower than that of the group fed with 100 lg
AFB1 kg)1 diet (Lim et al. 2001). This might due to the
dietary soybean meal was substituted by the aflatoxincontaminated palm kernel meal.

Aflatoxin B1 has been reported to impact the liver of
fish (Tuan et al. 2002). AST and ALT were used as the

..............................................................................................

Ó 2009 The Authors
Journal compilation Ó 2009 Blackwell Publishing Ltd Aquaculture Nutrition 16; 335–342


peroxidation or chromosomal damage through the release of
oxygen free radicals (Amstad et al. 1984; Shen et al. 1994),
which suggested that increasing oxidative damage might
contribute to the cytotoxic effects of AFB1. These findings
supported the results found in this study that fish consumed
more than 10 lg AFB1 kg)1 diet showed no alarming signs
except increasing serum SOD activities. On the other hand,
Rastogi et al. (2001) reported that SOD activity in liver of
rats was decreased when intraperitoneally administered at a
dose of 2 mg AFB1 kg)1 body weight.

Aflatoxin B1 in hepatopancreas
(AH, µg kg–1)

40
35
30
25

AH = 13.91 + 18.73 * log10 (AD)
R 2 = 0.8976, P < 0.05

20
15
10
5
0
0

5
10
15
20
25
Aflatoxin B1 in diets (AD, µg kg–1)

30

Aflatoxin B1 residue in tissues (µg kg–1)

Figure 2 The relationship between AFB1 in hepatopancreas (AH,
lg kg)1) and in diets (AD, lg kg)1).

7

6
5

Muscle
Ovary

4
3
2

In histological studies of AFB1, the liver of fish was one of
the organs most affected by dietary AFB1 (Hendricks 1994).
However, in this study, no significant changes were observed
in HSI and hepatopancreatic histology of gibel carp among
different groups, which suggested the low dietary AFB1 could
not cause the hepatopancreatic histological changes of gibel
carp. This was supported by the effects of low dietary AFB1
as 0–250 lg kg)1 on Nile tilapia (Tuan et al. 2002) and
0–100 lg kg)1 on black tiger shrimp (Boonyaratpalin et al.
2001), but a concentration as low as 0.5 lg of AFB1 kg)1 diet
fed for 6 months had been found to cause liver tumours in
rainbow trout (Halver 1969). Our results indicated that gibel
carp is much less sensitive than rainbow trout to the histopathological effects of AFB1 (Halver 1969).

1
0

5

10

15
20
25
Dietary aflatoxin B1 (µg kg–1)

30

Figure 3 AFB1 residue in muscle and ovary of gibel carp fed with
different dietary AFB1 for 3 months, and AFB1 residue in tissues of
fish was not detected in the control group and Diet 2.

biochemical indicators for hepatic damage (Cheng et al.
2001; Allameh et al. 2005). In this study, the increase in ALT
activity of fish fed with the highest dietary AFB1 was statistically significant, which suggested that AFB1 slightly
damaged the hepatopancreas of gibel carp. This result agreed
with the previous finding in broilers (Bintvihok &
Kositcharoenkul 2006). Similarly, Han et al. (2008) also
described a marked increase by 9.6% or 13.8% in serum
ALT activities of ducks fed the diets containing 20 lg kg)1
or 40 lg kg)1 AFB1-contaminated rice when compared to
the value of the control group free of AFB1.
The increase in activation of SOD conferred protection
against oxidative damage (Mattson et al. 1995). Many
studies have observed that AFB1 was able to induce lipid

1

A number of studies have reported analysis of aflatoxin
residues in cattle (Cook et al. 1986), broilers (Bintvihok &
Kositcharoenkul 2006) and lambs (Ferna´ndez et al. 1997)

but not in aquatic animal species. Ngethe et al. (1992)
reported that the highest AFB1 disposition was observed in
the bile, liver, kidney and pyloric caeca of rainbow trout
following oral and intravenous administration over a period
of 8 days. The finding was in close agreement with this study
that AFB1 residue in hepatopancreas was much higher than
that in muscles and ovaries. On the other hand, Divakaran &
Tacon (2000) observed that Aflatoxin B1 was not detected in
faeces, whole shrimp or tail muscle of Penaeus vannamei
when fed with dietary afaltoxin B1 (300, 400 and 900 lg
AFB1 kg)1) in 8 weeks.
Plakas et al. (1991) found that AFB1 residue in tissues
were rapidly depleted and there was a very low potential for
the accumulation of AFB1 and its metabolites in the edible
flesh of channel catfish through the consumption of AFB1contaminated feed. However, concern of indirect AFB1

..............................................................................................

Ó 2009 The Authors
Journal compilation Ó 2009 Blackwell Publishing Ltd Aquaculture Nutrition 16; 335–342


1

hazard of secondary contamination from eating fish flesh for
human has been raised (Madhusudhanan et al. 2006). In
European Union, the regulatory level for AFB1 in foods
adopted is 2 ppb posed on risk assessments (Food and
Agriculture Organization (FAO) 2004). In this study, AFB1
residue in muscles and ovaries of fish fed with more than

10 lg kg)1 diet was more than 2 ppb.

The growth hormesis response at low AFB1 levels of the
present study indicated that the effect of AFB1 could not be
effectively observed from the growth of fish, especially at low
concentrations. Fish fed with more than 10 lg AFB1 kg)1
diet showed impaired physiological responses and more
AFB1 residue of muscles and ovaries above the safety limitation of European Union.

The authors are grateful Guanghan Nie for his technical
help. This study was supported by National Key Technology
R&D Program (2007BAD37B02, 2001BA505B06) and partly
by Natural Science Foundation of China (30700626) and
Key Project of Hubei Provincial Science and Technology
Department (2006AA203A01).

Aebi, H. (1984) Catalase in vitro. Methods Enzymol., 105, 121–126.
Allameh, A., Safamehr, A., Mirhadi, S.A., Shivazad, M., RazzaghiAbyaneh, M. & Afshar-Naderi, A. (2005) Evaluation of
biochemical and production parameters of broiler chicks fed
ammonia treated aflatoxin contaminated maize grains. Anim. Feed
Sci. Technol., 122, 289–301.
Amstad, P., Levy, A., Emerit, I. & Cerutti, P. (1984) Evidence for
membrane-mediated chromosomal damage by AFB1 in human
lymphocytes. Carcinogenesis, 5, 719–723.
Belz, R.G., Cedergreen, N. & Sørensen, H. (2008) Hormesis in
mixtures – Can it be predicted? Sci. Total Environ., 404, 77–87.
Bintvihok, A. & Kositcharoenkul, S. (2006) Effect of dietary calcium
propionate on performance, hepatic enzyme activities and aflatoxin residues in broilers fed a diet containing low levels of aflatoxin B1. Toxicon, 47, 41–46.
Boonyaratpalin, M., Supamattaya, K., Verakunpiriya, V. &
Suprasert, D. (2001) Effects of aflatoxin B1 on growth performance, blood components, immune function and histopathological changes in black tiger shrimp (Penaeus monodon Fabricius).

Aquac. Res., 32(s1), 388–398.
Calabrese, E.J. (2005) Paradigm lost, paradigm found: the re-emergence of hormesis as a fundamental dose response model in the
toxicological sciences. Environ. Pollut., 138, 379–411.
Calabrese, E.J. (2008) Hormesis: why it is important to toxicology
and toxicologists. Environ. Toxicol. Chem., 27, 1451–1474.

Calabrese, E.J. & Baldwin, L.A. (1998) Hormesis as a biological
hypothesis. Environ. Health Perspect., 106, 357–362.
Calabrese, E.J. & Baldwin, L.A. (2003) The hormetic dose–response
model is more common than the threshold model in toxicology.
Toxicol. Sci., 71, 246–250.
Cha´vez-Sa´nchez, M.C., Martı´ nez Palacios, C.A. & Osorio Moreno,
I. (1994) Pathological effects of feeding young Oreochromis niloticus diets supplemented with different levels of aflatoxin B1.
Aquaculture, 127, 49–60.
Cheng, Y.-H., Shen, T.-F., Fei Pang, V. & Chen, B.-J. (2001) Effects
of aflatoxin and carotenoids on growth performance and immune
response in mule ducklings. Comp. Biochem. Physiol., 128C, 19–26.
Cook, W.O., Richard, J.L., Osweiller, G.D. & Trampell, D.W.
(1986) Clinical and pathologic changes in acute aflatoxicosis:
rumen motility and tissue and fluid concentrations of aflatoxins B1
and M1. Am. J. Vet. Res., 47, 1817–1825.
Diaz, G.J., Calabrese, E. & Blain, R. (2008) Aflatoxicosis in chickens
(Gallus gallus): an example of hormesis? Poult. Sci., 87, 727–732.
Divakaran, S. & Tacon, A.G.J. (2000) Studies on the toxicity of
shrimp (Penaeus vannamei) fed diets dosed with aflatoxin B1 to
humans. J. Aquat. Food Prod. Technol., 9, 115–120.
van Eijkeren, J.C., Bakker, M.I. & Zeilmaker, M.J. (2006) A simple
steady-state model for carry-over of aflatoxins from feed to cowÕs
milk. Food Addit. Contam., 23, 833–838.
Food and Agriculture Organization (FAO) (2004) Worldwide regulations for mycotoxins in food and feed in 2003. Food and

Nutrition Paper No. 81. FAO, Rome, Italy.
Food and Drug Administration (FDA) (1989) Action levels for
aflatoxins in animal feeds. FDA Compliance Policy Guide
7126.33. Office of Enforcement, Division of Compliance Policy,
1989, FDA, Washington, D.C., 3 pp.
Ferna´ndez, A., Belı´ o, R., Ramos, J.J., Sanz, M.C. & Sa´ez, T. (1997)
Aflatoxins and their metabolites in the tissues, faeces and urine
from lambs feeding on an aflatoxin-contaminated diet. J. Sci. Food
Agr., 74, 161–168.
Gowda, N.K.S., Suganthi, R.U., Malathi, V. & Raghavendra, A.
(2007) Efficacy of heat treatment and sun drying of aflatoxincontaminated feed for reducing the harmful biological effects in
sheep. Anim. Feed Sci. Tech., 133, 167–175.
Halver, J.E. (1969) Aflatoxicosis and Trout hepatoma. In: Aflatoxin
(Goldblatt, L.A. ed), pp. 265–306. Academic Press, New York.
Han, X.Y., Huang, Q.C., Li, W.F., Jiang, J.F. & Xu, Z.R. (2008)
Changes in growth performance, digestive enzyme activities and
nutrient digestibility of cherry valley ducks in response to aflatoxin
B1 levels. Livest. Sci., 119, 216–220.
Hendricks, J.D. (1994) Carcinogenicity of aflatoxins in nonmammalian organisms. In: Toxicology of Aflatoxins: Human
Health, Veterinary, and Agricultural Significance (Eaton, D.L. &
Groopman, J.D. eds), pp. 103–136. Academic Press, San Diego.
Hendricks, J.D., Putnam, T.P., Bills, D.D. & Sinnhuber, R.O. (1977)
Inhibitory effect of a polychlorinated biphenyl (Aroclor 1254) on
aflatoxin B1 carcinogenesis in rainbow trout (Salmo gairdneri).
J. Natl Cancer Inst., 59, 1545–1551.
Hendricks, J.D., Meyers, T.R., Shelton, D.W., Castell, J.L. & Bailey,
G.S. (1985) Hepacarcinogenicity of benzo(a)pyrene to rainbow
trout by dietary exposure and intraperitoneal injection. J. Natl
Cancer Inst., 74, 839–851.
Hørder, H. & Rej, R. (1981) Aminotransferase. In: Methods of

Enzymatic Analysis, vol. 3 (Bergmeyer, H.U. Ed.), pp. 416–424,
444–450. Gmbh press, Verlag Chemie.
Jantrarotai, W. & Lovell, R.T. (1990) Subchronic toxicity of dietary
aflatoxin B1 to channel catfish. J. Aquat. Anim. Health, 2, 248–254.

..............................................................................................

Ó 2009 The Authors
Journal compilation Ó 2009 Blackwell Publishing Ltd Aquaculture Nutrition 16; 335–342


Johnson, T.E. & Bruunsgaard, H. (1998) Implications of hormesis
for biomedical aging research. Hum. Exp. Toxicol., 17, 263–265.
Lim, H.-A., Ng, W.-K., Lim, S.-L. & Ibrahim, C.O. (2001) Contamination of palm kernel meal with Aspergillus flavus affects its
nutritive value in pelleted feed for tilapia, Oreochromis mossambicus. Aquac. Res., 32, 895–905.
Lovell, R.T. (1989) Nonnutrient diet components. In: Nutrition and
Feedings of Fish (Lovell, T. ed.), p. 93. Van Nostrand Reinhold,
New York, N.Y.
Madhusudhanan, N., Kavithalakshmi, S.N., Shanmugasundaram,
E.R. & Shanmugasundaram, K.R. (2006) Aflatoxin B1-induced
DNA damage in Labeo rohita: protective effect of an antioxidant
supplement, Amrita Bindu. Basic Clin. Pharmacol. Toxicol., 98,
473–479.
Madrigal-Santilla´n, E., Madrigal-Bujaidar, E., Ma´rquez-Ma´rquez,
R. & Reyes, A. (2006) Antigenotoxic effect of Saccharomyces
cerevisiae on the damage produced in mice fed with aflatoxin B1
contaminated corn. Food Chem. Toxicol., 44, 2058–2063.
Mattson, M.P., Lovell, M.A., Furukawa, K. & Markesbery, W.R.
(1995) Neurotrophic factors attenuate glutamate-induced accumulation of peroxides, elevation of intracellular Ca2+concentration, and neurotoxicity and increase antioxidant enzyme activities
in hippocampal neurons. J. Neurochem., 65, 1740–1751.

McCord, J.M. & Fridovich, I. (1969) Superoxide dismutase: an
enzymatic function for erythrocuprein (hemocuprein). J. Biol.
Chem., 244, 6049–6055.
McKean, C., Tang, L., Tang, M., Billam, M., Wang, Z., Theodorakis, C.W., Kendall, R.J. & Wang, J.S. (2006) Comparative acute
and combinative toxicity of aflatoxin B1 and fumonisin B1 in
animals and human cells. Food Chem. Toxicol., 44, 868–876.
Murado, M.A. & Va´zquez, J.A. (2007) The notion of hormesis and
the dose–response theory: A unified approach. J. Theor. Biol., 244,
489–499.
Ngethe, S., Horsberg, T.E. & Ingebrigtsen, K. (1992) The disposition
of 3H-aflatoxin B1 in the rainbow trout (Oncorhynchus mykiss)

after oral and intravenous administration. Aquaculture, 108, 323–
332.
Ottinger, C.A. & Kaattari, S.L. (1998) Sensitivity of rainbow trout
leucocytes to aflatoxin B1. Fish Shellfish Immunol., 8, 515–530.
Plakas, S.M., Loveland, P.M., Bailey, G.S., Blazer, V.S. & Wilson,
G.L. (1991) Tissue disposition and excretion of 14C-labelled aflatoxin B1 after oral administration in channel catfish. Food Chem.
Toxicol., 29, 805–808.
Rastogi, R., Srivastava, A.K. & Rastogi, A.K. (2001) Biochemical
changes induced in liver and serum of aflatoxin B1-treated male
wistar rats: preventive effect of picroliv. Pharmacol. Toxicol., 88,
53–58.
Sahoo, P.K. & Mukherjee, S.C. (2001) Effect of dietary – 1,3 glucan
on immune responses and disease resistance of healthy and
aflatoxin B1-induced immunocompromised rohu (Labeo rohita
Hamilton). Fish Shellfish Immunol., 11, 683–695.
Sahoo, P.K. & Mukherjee, S.C. (2003) Immunomodulation by
dietary vitamin C in healthy and aflatoxin B1-induced immunocompromised rohu (Labeo rohita). Comp. Immunol. Microbiol.
Infect. Dis., 26, 65–76.

Sahoo, P.K., Chattopadhyay, S.K. & Sikdar, A. (1996) Immunosuppressive effects of induced aflatoxicosis in rabbits. J. Appl.
Anim. Res., 9, 17–26.
Santacroce, M.P., Conversano, M.C., Casalino, E., Lai, O.,
Zizzadoro, C., Centoducati, G. & Crescenzo, G. (2008) Aflatoxins
in aquatic species: metabolism, toxicity and perspectives. Rev. Fish
Biol. Fisheries, 18, 99–130.
Shen, H.M., Shi, C.Y., Lee, H.P. & Ong, C.N. (1994) Aflatoxin B1induced lipid peroxidation in rat liver. Toxicol. Appl. Pharmacol.,
127, l45–l150.
Tuan, N.A., Grizzle, J.M., Lovell, R.T., Manning, B.B. & Rottinghaus, G.E. (2002) Growth and hepatic lesions of Nile tilapia
(Oreochromis niloticus) fed diets containing aflatoxin B1. Aquaculture, 212, 311–319.

..............................................................................................

Ó 2009 The Authors
Journal compilation Ó 2009 Blackwell Publishing Ltd Aquaculture Nutrition 16; 335–342


Aquaculture Nutrition
2010 16; 343–350

..........................................................................................

1,2
1

doi: 10.1111/j.1365-2095.2009.00670.x

1

Department of Fisheries – Research Division, North Beach, WA, Australia; 2 Present address, CSIRO Marine and Atmospheric

Research, Cleveland, Queensland, Australia

Several dietary strategies to ameliorate poorer growth
observed to occur at temperatures above the upper thermal
optima were examined with juvenile barramundi (Lates
calcarifer). A reference (REF) and three experimental diets,
one with an increased protein to energy ratio (PRO), another
with an increased level of the amino acid histidine (HIS) and
a third with supplementation of dietary nucleotides (NUC),
were each fed to fish at either 30 °C or 37 °C for a 28-day
period. Growth was affected by both temperature and diet.
Fish fed the PRO diet at 30 °C grew fastest, but not faster
than those fed the NUC diet at the same temperature. The
addition of the amino acid histidine to the diet did not
improve growth rates at either temperature. At water temperatures of 37 °C, only the fish fed the PRO diet had growth
rates equivalent to those of fish at the 30 °C temperatures.
Other key factors including feed intake, feed conversion rate,
nutrient and energy retention and plasma enzymology were
also all affected by temperature and diet. This study shows
that the use of a diet with an increased protein to energy ratio
provides significant benefits in terms of reducing the impact
of growth retardation at higher temperatures.
KEY WORDS:

Asian seabass, heat, stress, temperature

Received 26 September 2008, accepted 4 March 2009
Correspondence: Brett Glencross, CSIRO Marine and Atmospheric
Research, PO Box 120, Cleveland, QLD 4163, Australia. E-mail: Brett.

Barramundi production in Australia occurs in water temperatures that range from <20 °C to 35 °C (Boonyaratpalin &
Williams 2001; Glencross 2006). However, optimal growth

..............................................................................................

Ó 2009 The Authors
Journal compilation Ó 2009 Blackwell Publishing Ltd

has been shown to occur for this species at around 30 °C and
that above 33 °C the growth and feed utilization performance
of the fish begins to decline as the fish start to experience heat
stress (Katersky & Carter 2005). Dietary strategies to improve
performance for this species have been examined at the lower
temperature spectrum, but little work has been performed at
upper physiological ranges (Williams et al. 2006).
Temperature stress manifests several key physiological
responses in fish, growth retardation and death being some of
the extreme outcomes. However, other key processes that
occur include the development of cataracts, a reduction in
protein digestibility, an increase in endogenous protein
turnover and a decline in endogenous protein synthesis
(Carter et al. 2006; M. Bermudes, personal communication).
The increased rate of protein turnover during heat stress
suggests that an increase in the dietary protein to energy ratio
might also provide a means of improving the fishÕs response
to this higher level of protein turnover (Barnes et al. 2006).
However, studies with Atlantic salmon found no significant
benefit from changes to the dietary protein to energy ratio,
although it was acknowledged that the range of dietary

protein to energy ratios was not broad.
With the higher turnover rate of protein, there is likely to
be an increased demand for amino acids in general and
potentially some specific ones more than others (Carter et al.
2006). The observation of cataracts among barramundi
under heat stress draws analogies with other pathological
causes of cataracts, including a dietary deficiency of the
amino acid histidine (Breck et al. 2005).
Another dietary supplement that has also been used to
support fish during periods of stress is nucleotides (Li &
Gatlin 2006). Nucleotides have been regarded as conditional
or semi-essential nutrients for fish. Their inclusion in fish diets
has been shown to enhance immunity and disease resistance
of fish by influencing the immune system. Nucleotide


supplementation has also been shown to affect the growth and
differentiation of cells from the gastrointestinal tract of fish
(Uauy et al. 1990), an organ that is critically affected by
increased levels of protein turnover during heat stress.
This study examines a series of diets and the potential
benefits they provide to barramundi, being grown at temperatures higher than their optimal range, compared with a
standard reference diet and their optimal growth temperature. Each of the dietary treatments considered were included
based on their potential to address key effects of heat stress
previously identified. A diet with an increased protein to
energy ratio (PRO) was included to see if this could compensate for the increased rate of protein turnover and decline
in protein digestibility. A diet fortified with histidine (HIS)
was included because of the reported benefits of this dietary
amino acid in reducing incidence of cataracts (Breck et al.
2005). A third diet included an allocation of nucleotides

(NUC), which have been implicated in improving the ability
of fish to deal with stress (Li & Gatlin 2006).

Each of the four experimental diets was extruded using an
APV 19:25 laboratory-scale twin-screw feed extruder through
a 2.4 mm diameter die. Following extrusion, the pellets were
oven-dried at 60 °C for approximately 12 h. Following drying, and while the pellets were still warm, they were vacuuminfused with the formulated oil allocation. The diet complete
formulations and source of all of the ingredients used is
presented in Table 1. Composition of each of the experimental diets is also presented in Table 1.

This study used hatchery-reared barramundi (Lates calcarifer)
that were pregrown to an acclimation weight of around 15 g in
indoor 1000-L heated sea-water tanks. For acclimation the
fish were split into two groups of four 150-L tanks of around
100 fish each. One group of four tanks was maintained at
30 °C consistent with the original temperature at which the
fish were grown. The other group of four tanks had their water
temperature increased by 2 °C every 72 h over an 8-day period
to a maximum temperature of 37 °C to minimize acclimation
stress (Mora & Maya 2006). Following the acclimation
period, 15 fish were each randomly allocated to each of the
12 tanks within their respective temperature blocks with an
overall initial weight of 17.9 ± 0.25 (mean ± SD, n = 360),

Table 1 Nutrient composition of the experimental diets
PRO

NUC

HIS

REF

1
5
65
60
84
0
785
0

1
5
150
0
142
0
700
2

1
5
150
0
124
20
700
0

1
5
150
0
144
0
700
0

949
599
543

950
518
441

951
530
441

928
523
441

145
114
143
22.2
20.2

202
158
123
23.5
21.1

209
134
127
23.6
20.9

210
137
131
23.4
21.1

26.9
617
38
17
28
51
41
21
29
29
28

20.9
480
30
14
23
40
35
17
23
23
23

21.1
537
31
34
24
43
46
19
24
24
24

20.9
514
31
16
23

42
43
18
23
24
24

)1

Ingredients (g kg )
Ytterbium oxide1
Pre-mix vitamins2,7
Fish oil3
Wheat gluten4
Wheat flour4
5
L-histidine
3
Fish meal
Nucleotides6
Composition – all
values in g kg)1
DM unless otherwise
detailed
DM
Protein
Estimated digestible
protein8
Lipid
Carbohydrate

Ash
Gross energy (MJ kg)1 DM)
Estimated digestible energy
(MJ kg)1 DM)9
DP–DE (g MJ)1)10
Sum of amino acids
Arginine
Histidine
Isoleucine
Leucine
Lysine
Methionine
Phenylalanine
Threonine
Valine

1
Ytterbium oxide sourced from Stanford Materials, Aliso Viejo, CA,
USA.
2
Sourced from Adisseo Animal Nutrition, Goodna, Queensland,
Australia.
3
Sourced from Skretting Australia, Cambridge, Tasmania, Australia.
4
Sourced from Manildra, Auburn, New South Wales, Australia.
5
Sourced from Sigma, St Louis, Missouri, USA.
6
Sourced from Ridley Aquafeeds, Narangba, Queensland, Australia, as OptimunTM.

7
Vitamin and mineral premix includes (IU kg)1 or g kg)1 of premix): vitamin A, 2.5 mIU; vitamin D3, 0.25 mIU; vitamin E, 16.7 g;
vitamin K3, 1.7 g; vitamin B1, 2.5 g; vitamin B2, 4.2 g; vitamin B3,
25 g; vitamin B5, 8.3 g; vitamin B6, 2.0 g; vitamin B9, 0.8 g; vitamin
B12, 0.005 g; biotin, 0.17 g; vitamin C, 75 g; choline, 166.7 g; inositol, 58.3 g; ethoxyquin, 20.8 g; copper, 2.5 g; ferrous iron, 10.0 g;
magnesium, 16.6 g; manganese, 15.0 g; zinc, 25.0 g.
8
Based on protein digestibility of wheat gluten at 100%, wheat
flour at 100% and fishmeal at 90%.
9
Based on energy digestibility of fish oil at 95%, wheat gluten at
95%, wheat flour at 50% and fishmeal at 95%.
10
DP–DE: digestible protein to digestible energy ratio. Digestible
protein and energy values are derived from McMeniman 1998.
PRO, high protein to energy ratio diet; NUC, nucleotide fortified
diet; HIS, histidine fortified diet; REF, reference experimental diet;
DM, dry matter.

..............................................................................................

Ó 2009 The Authors
Journal compilation Ó 2009 Blackwell Publishing Ltd Aquaculture Nutrition 16; 343–350


with initial block weights of 17.9 ± 0.24 for the 30 °C temperature block and 17.8 ± 0.26 for the 37 °C temperature
block. Each tank was supplied with flow-through (4 L min)1),
heated [temperature controlled by mixing of 20 °C and 40 °C
water, controlled by a digital programmable logic controllermanaged solenoid and water mixing box prior to the water
being distributed to each tank] and continuous aerated seawater. Over the course of the study, the temperatures for each

of the two temperature blocks were 29.8 ± 0.3 °C and
36.6 ± 0.4 °C. Prior to weighing, the fish were sedated using
isoeugenol at 0.002 mL L)1 (supplied as AQUI-STM, AQUI-S
Pty Ltd, Lower Hutt, New Zealand) and individually weighed
to 0.1 g accuracy. The fish were then allowed to regain consciousness and equilibrium before being placed within their
designated tank. At the end of the study, the fish were again
sedated and weighed as detailed previously.

The fish were fed to apparent satiety once daily between 09:00
and 11:00 over a 28-day (4-week) period. Apparent satiety was
determined by a loss in feeding activity, this was reached after
three feeding sessions over a 1-h period. Any uneaten feed was
removed from each tank 1 h later and the uneaten portion was
dried and weighed to allow the determination of daily feed
intake based on correction factors for leaching losses sustained over an equivalent period (Helland et al. 1996).

At the end of the experiment, after the final weighing, three fish
from each tank were pooled and processed for sample analysis.
The three fish were minced together after being dried of
residual surface moisture. Samples of the mince were then
taken for dry matter analysis and another was frozen prior to
being freeze-dried. Freeze-dried whole fish samples were
analysed for dry matter, nitrogen, ash, fat, amino acids and
energy. Blood samples were collected from an additional three
fish from each tank using a 1-mL syringe and an 18G needle via
the caudal vein. Blood from each of the three fish was pooled
within a single Eppendorf TM (Eppendorf, Sydney, Australia)
tube, centrifuged at 1000g for 1 min to settle the erythrocytes
and the plasma transferred to another Eppendorf TM tube
prior to it being frozen and sent for plasma analysis.

All chemical analyses were performed by NATA (National
Association of Testing Authorities) accredited analytical

service providers (Chemistry Centre, East Perth, WA,
Australia). Diet and fish samples were analysed for dry
matter, ash, nitrogen, total lipids, amino acids and gross
energy content. Dry matter was calculated by gravimetric
analysis following oven drying at 105 °C for 24 h. Protein
levels were calculated from the determination of total nitrogen by Leco auto-analyser, based on N · 6.25. Amino acid
composition of samples was determined by an acid hydrolysis
prior to separation via high-performance liquid chromatography. The acid hydrolysis destroyed tryptophan making it
unable to be determined using this method. Crude fat content
of the diets was determined gravimetrically following
extraction of the lipids according to the method of Folch
et al. (1957). Gross ash content was determined gravimetrically following loss of mass after combustion of a sample in a
muffle furnace at 550 °C for 12 h. Gross energy was determined by adiabatic bomb calorimetry.
Samples of plasma were sent to the West Australian Animal
Health Laboratories (South Perth, WA) for plasma enzyme
assessment. The assays were run on an Olympus AU400
automated chemistry analyser (Olympus Optical Co. Ltd.,
Ngano, Japan). Each of the assays used was a standard kit
developed for the auto-analyser. The tests performed included
alanine aminotransferance (ALT; Olympus Kit Cat. No.
OSR6107), direct bilirubin (Olympus Kit Cat. No. OSR6111),
total bilirubin (Olympus Kit Cat. No. OSR6112), creatine
kinase (CK; Olympus Kit Cat. No. OSR6179), gamma-glutamyltransferase (GGT; Olympus Kit Cat. No. OSR6219),
glutamate dehydrongenase (GLDH; Randox Kit Cat. No.
GL441), lactate dehydrogenase (LDH; Olympus Kit Cat.
No. OSR6128) and total protein (Olympus Kit Cat. No.

OSR6132).

Nitrogen, energy, lysine and histidine retention were determined based on the mass gain of each parameter over the
course of the growth study, against the respective dietary intake of each respective parameter. All values were calculated
according to the following formula (Maynard & Loosli 1969):
Nitrogen retention (%) ¼



Nt À Ni
 100;
Nc

where Nt is the nutrient/energy content of the fish in a specific replicate at time t and Ni is the mean initial nutrient/
energy content of the fish from the beginning of the study
(n = 3 replicates of three representative fish); Nc is the
amount of nutrient/energy consumed by the fish from the

..............................................................................................

Ó 2009 The Authors
Journal compilation Ó 2009 Blackwell Publishing Ltd Aquaculture Nutrition 16; 343–350


time of initial assessment to time t. Determination of energy
and amino acid retentions was achieved the same way, but
with the substitution of the relevant energy or amino acid
criteria where the corresponding nitrogen criteria are indicated in the equation.

All values are means unless otherwise specified. All data

were analysed for homogeneity of variances using CochranÕs
test. Effects of diet, temperature and interactions were
analysed using a two-factor MANOVA using the software
package Statistica (StatsoftÒ, Tulsa, OA, USA). Significance
of differences were further tested using a FisherÕs least significant difference test based on critical ranges being set at
P < 0.05.

Significant effects of both diet and temperature were
observed on the final weight and weight gain of the fish
(Table 2). No interaction effect between diet and temperature
was observed for the final weight and weight gain. Fish at
30 °C all grew better than those fed the same diet at 37 °C,
the only exception being the PRO diet, with which the difference in growth between the two temperatures was not
significant. At 30 °C temperatures, the fish fed the PRO and
NUC diets grew significantly better than those fed the reference (REF) diet, but not the HIS diet. At 37 °C temperatures, the fish fed the PRO diet grew significantly better than
those fed the NUC, HIS and REF diets.

No significant effects of either diet and temperature, or
interaction terms were observed on feed intake by the fish
(Table 2). Post-hoc analysis did identify that fish fed the
NUC diet at 37 °C ate less feed than all the other treatments.
However, significant effects of diet, temperature and the
interaction terms were observed on the feed conversion rate
(FCR; Table 2). Fish at 30 °C all had lower FCR than those
fed the same diet at 37 °C. At 30 °C temperatures, the fish
fed the PRO and NUC diets had significantly better FCR
than those fed the HIS or REF diet, but there were no differences in the FCR between the HIS or REF diets. At 37 °C
temperatures, the fish fed the PRO and NUC diets had significantly lower FCR than those fed the HIS and REF diets,
while the FCR of the HIS diet was also lower than that of the
REF diet.

There were no significant effects of diet or interaction term
on energy retention efficiency, but there was a significant
effect of temperature (Table 2). Energy retention was significantly poorer at 37 °C than 30 °C. Within the 30 °C temperature, both the PRO and NUC diets had better energy
retention than both the HIS and the REF diets. Within the
37 °C temperature, there were no significant differences in
energy retention efficiency. There were no significant effects
of diet or interaction term on protein retention efficiency, but
there was a significant effect of temperature (Table 2). Within
the 30 °C temperature, both the PRO and NUC diets had
better protein retention than the HIS diet, but not the REF
diet. Within the 37 °C temperature, there were no significant
differences in protein retention efficiency.
The retention of specific amino acids like lysine and histidine was significantly affected by the diet and interaction
terms, but was not affected by temperature (Table 2). At
30 °C, the retention of both lysine and histidine was highest

Table 2 Growth performance and feed utilization by fish fed the experimental diets

Diet

Units

PRO

NUC

HIS

REF

PRO

NUC

HIS

REF

P-value
Pooled
SEM
Diet Temperature T · D

Initial weight
Final weight
Gain
FCR
Intake
Energy retention
Protein retention
Lysine retention
Histidine retention
Fat retention

g per fish
g per fish
g per fish
Feed per gain
g per fish
%

%
%
%
%

17.9
61.7a
43.7a
0.65a
28.3a
63.0a
63.2a
82.9a
40.0a
57.0ab

18.1
61.5a
43.5a
0.65a
28.4a
63.0a
61.3a
46.6b
37.0a
60.4a

17.9
56.6ab
38.6ab

0.72b
27.7a
50.9b
48.9bc
21.7d
9.6d
50.8bc

17.9
53.1b
35.2b
0.77bc
27.0ab
51.4b
53.6ab
21.4d
19.7c
43.2c

17.8
57.6a
39.8ab
0.72b
28.7a
46.2b
41.5bc
48.9b
30.2b
64.4a

17.9
48.6c
30.7c
0.75b
22.9b
50.8b
49.5bc
47.5b
38.1a
50.4bc

17.6
49.9bc
32.4bc
0.82c
26.7ab
48.9b
43.7bc
29.4cd
15.0c
50.7 bc

18.0
47.7c
29.7c
0.98d
29.2a
44.8b
41.0c
35.1c

27.7b
45.3c

0.05
1.32
1.31
0.02
0.64
1.70
2.00
4.00
2.27
1.83

30 °C

Temperature

37 °C

–
0.016
0.016
0.000
0.393
0.068
0.113
0.000
0.000
0.004

–
0.001
0.001
0.000
0.431
0.001
0.000
0.177
0.341
0.951

–
0.347
0.349
0.004
0.195
0.185
0.266
0.000
0.000
0.014

Different superscripts indicate significant differences (P < 0.05) among treatments. P-value (significance values) for temperature, diet and
interaction effects determined using two-way ANOVA (MANOVA) with FisherÕs least significant difference post-hoc tests.
PRO, high protein to energy ratio diet; NUC, nucleotide fortified diet; HIS, histidine fortified diet; REF, reference experimental diet.

..............................................................................................

Ó 2009 The Authors

Journal compilation Ó 2009 Blackwell Publishing Ltd Aquaculture Nutrition 16; 343–350


in fish fed the PRO diet and lowest in fish fed the HIS or
REF diets. However, at 37 °C there was less distinction
between diets in terms of lysine and histidine retention efficiencies, although similar to that observed at 30 °C the
lowest retention efficiency was that of histidine from the HIS
diet and the highest was that of lysine from the PRO diet.
Lipid retention was affected by the diet and interaction
terms, but there was no significant effect of temperature
(Table 2).

A series of significant effects on the plasma chemistry were
attributable to temperature, but no diet or interaction term
effects were noted (Table 3). Some discrete differences were
however observed at each temperature for specific parameters.
At 30 °C temperatures, the fish fed the PRO, NUC and
HIS diets had significantly lower CK activity than those fed
the REF diet. At 37 °C temperatures, the fish fed the NUC
diet had significantly higher CK than those fed the PRO, HIS
and REF diets.
Within the 30 °C temperatures, the fish fed the PRO diet
had significantly lower ALT activity than those fed the NUC,
HIS and REF diets. At 37 °C temperatures, the fish fed the
NUC diet had significantly higher CK than those fed the
PRO, HIS and REF diets.
Of the fish maintained at 30 °C temperatures, those fed the
HIS and REF diets had significantly higher GLDH activities
than those fed the PRO and NUC diets. At 37 °C temperatures, the fish fed the NUC diet had significantly higher
GLDH than those fed the PRO, HIS and REF diets.

LDH activity also varied among treatments within temperatures. At 37 °C temperatures, the fish fed the NUC diet
had significantly higher LDH than those fed the PRO, HIS
and REF diets. However, there were no significant differences among the fish maintained at the 30 °C temperatures.

Based on the MANOVA there were no significant differences
of treatment, diet or interaction term. Although total protein
levels in the plasma showed no significant variability among
diets at 30 °C, those fish at 37 °C fed the NUC diet had
significantly lower plasma protein levels than those fed any of
the other diets.

This study examined the potential of a range of dietary
strategies to minimize the impact of elevated water temperatures on performance parameters of barramundi. Clear
effects of temperature, diet and some diet · temperature
interaction terms were identified. These observations provide
a clear indication that dietary strategies can be used to reduce
the effects of heat stress on growth. In addition, clear benefits
of certain dietary strategies on growth even at optimal temperatures were also identified.

Elevated temperature had a clear effect on all growth
parameters, but not so on feed intake. This observation is
consistent with those reported by Katersky & Carter (2005)
who also observed a reduction in growth of fish above 35 °C
relative to that at 30 °C, but not a commensurate reduction
in feed intake. In a similar study with Atlantic salmon, no
benefit was observed from increasing the dietary protein to
energy ratio (Barnes et al. 2006). However, in contrast to the
present study the range of dietary protein to energy ratio
examined in the study by Barnes et al. (2006) was not as great
and this may have not allowed any significant effects to have

been observed. Consistent with the maintenance of feed
intake levels, but with poorer growth the feed conversion of
fish from the 37 °C temperature series was also significantly
poorer, similar to the observations of Katersky & Carter

Table 3 Plasma protein and enzyme activities from barramundi fed each of the dietary treatments
30 °C

Temperature
Diet
Creatinine kinase
Alanine aminotransferase
Glutamate dehydrongenase
Lactate dehydrogenase
Total plasma protein

Units PRO
)1

37 °C
NUC

a

HIS
ab

REF
a

PRO
b

NUC
c

HIS
d

REF
c

P-value
Pooled
SEM
Diet Temperature T · D

UL
9886 10 337 9252 16 360 29 073 46 767 30 777 29 542c 3299
)1
UL
16a
35b 25ab
33b
56c
68d
48c
48c
5
U L)1

14a
14a
20b
24b
29bc
46d
34c
33c
3
U L)1 3428a
4130a 4766a
5060a 11 033b 16 950c 10 594b 10 927b 1088
g L)1
42a
43ab 44ab
41a
42ab
33c
45b
40a
1

0.535
0.667
0.344
0.343
0.238

0.000
0.016

0.000
0.000
0.315

0.391
0.831
0.127
0.259
0.257

Different superscripts indicate significant differences (P < 0.05) among treatments. P-value (significance values) for temperature, diet and
interaction effects determined using two-way ANOVA (MANOVA) with FisherÕs least significant difference post-hoc tests.
PRO, high protein to energy ratio diet; NUC, nucleotide fortified diet; HIS, histidine fortified diet; REF, reference experimental diet.

..............................................................................................

Ó 2009 The Authors
Journal compilation Ó 2009 Blackwell Publishing Ltd Aquaculture Nutrition 16; 343–350


(2005). Retention of energy and protein was also significantly
impaired at the elevated temperatures (consistent with the
poorer FCR), but based on the MANOVA analysis the retention of lysine and histidine were not significantly impaired by
elevated water temperatures, although certainly several of the
diets had lower lysine and histidine retention efficiencies at
37 °C compared with the same diets at 30 °C. Feed intake of
the NUC diet at 37 °C was significantly reduced and no
apparent benefit from the inclusion of nucleotides on any
of the growth parameters could be ascertained at this
temperature.

Several key dietary effects were observed in these studies that
have important implications for improving the performance
of barramundi diets. An increase in the protein to energy
ratio significantly improved growth of fish relative to the
REF diet, but similar gains were also observed from the
NUC diet. These improvements in performance associated
with a higher protein to energy ratio are consistent with
idealized diet specifications determined from modelling
studies with this species (Glencross 2008). Improvements in
growth with the addition of nucleotides to the diet also
provided significant benefit, equal to that achieved with the
increase in protein to energy ratio. These observations support that at optimal growing conditions, where nutrient
demand is highest, that nucleotides may indeed be regarded
as conditionally or semi-essential nutrients (Li & Gatlin
2006). Importantly, it was noted that the addition of nucleotides to the diet of fish at 30 °C also significantly improved
the retention efficiencies of energy, protein, lysine and histidine relative to the REF diet. Further work identifying the
specific nature of this improvement in protein metabolism
with the inclusion of nucleotides is certainly warranted.
A series of dietary effects on energy and nutrient retention
efficiencies were also observed. High protein to energy ratios
were commensurate with improved energy and protein
retention and high lysine retention efficiency at both temperatures, particularly so at 30 °C. Similarly, the inclusion of
nucleotides in the diet (NUC) also improved most retention
parameters compared with those observed in the REF diet.
The addition of crystalline histidine to the diet (HIS) resulted
in significantly reduced histidine retention efficiencies at both
temperatures, but did not affect energy, protein or lysine
retention. Comparison with the other treatments suggests
that this lower retention efficiency was as a result of a poorer

ability of the fish to use either elevated histidine levels or the

crystalline amino acids. In other studies examining the
retention of crystalline amino acids they have generally been
found to be well utilized when included at low levels,
equivalent to that of protein-bound amino acids (Williams
et al. 2001).

Key interaction terms were identified for the retention efficiencies of both lysine and histidine. The retention of lysine in
the PRO diet at 30 °C was very high, significantly more so
than that from the NUC diet fed fish. However at 37 °C, this
difference was not present. In contrast to that observed with
the other three diets, an increase in the retention efficiency of
both lysine and histidine at 37 °C from fish fed the REF diet
was also observed relative to that at 30 °C. This observation
is unusual in that the retention characteristics of these amino
acids are counter to that of the total dietary protein content.
The improved retention at 37 °C of these two amino acids
from the REF diet relative to that at 30 °C suggests that
some amino acids may be used more efficiently at higher
temperatures when present at a lower ratio to energy content,
while the total protein retention clearly shows no benefits,
which suggests that other amino acids must be used substantially less efficiently.
No other interaction effects were observed for any of the
growth parameter, supporting that growth was discretely
influenced by diet or temperature.

The use of blood plasma enzymology provides an additional
insight into the effects of each diet and temperature on the
biochemical processes occurring within the animal. Each of

the enzymes examined in this study are key marker enzymes
for particular cell types or metabolic pathways and as such
provide key information on the effects of each of the treatments on certain tissues and metabolic processes.
Elevated CK activities at the higher water temperatures are
indicative of broad-scale muscle cell damage (Kopp et al.
2004). Within the 30 °C temperature, reduced CK levels with
the PRO and HIS diets compared with the REF diet support
that these diets reduce the level of muscle cell damage. At
37 °C, these benefits of the PRO and HIS diets were not
present and the NUC diet had an even further elevated level
of CK activity compared with the REF diet. These observations support that at elevated temperatures the inclusion of
nucleotides may exacerbate muscle cell damage.

..............................................................................................

Ó 2009 The Authors
Journal compilation Ó 2009 Blackwell Publishing Ltd Aquaculture Nutrition 16; 343–350


A similar elevation in ALT activity at the elevated water
temperatures indicates an increase in liver cell damage (Begum
2004; Refstie et al. 2006; Saoudi et al. 2008). At optimal water
temperatures, a reduction in ALT activity was observed
compared with those at 37 °C. The increase of dietary protein
to energy ratio also reduced the level of ALT activity supporting that the PRO diet limited the incidence of liver damage
at optimal water temperatures, but not at elevated water
temperatures. The NUC diet while not exacerbating ALT
activity at optimal temperatures did increase at 37 °C relative
to the REF diet treatment at the same temperature.
The increase in GLDH activities at elevated water temperatures are consistent with the increased level of endogenous protein turnover as this enzyme is involved in the

breakdown of amino acids by the reduction of glutamate to
a-ketoglutarate releasing ammonia and allowing access of
amino acids to the tricarboxylic acid cycle (Begum 2004). The
enzyme is also regarded as a marker enzyme of non-muscle
tissue (e.g. liver and kidney) and therefore elevated levels
support that the tissue damage occurring is non-muscle
damage. That the PRO and NUC diets had further reduced
GLDH activities at 30 °C supports that these diets accentuated protein retention and this is consistent with the observed
protein and amino acid retention data.
LDH activities are generally indicative of tissue degeneration (Rao & Venkateswara 2006). The elevated levels of
LDH at 37 °C are indicative of an increased level of tissue
damage and are also supportive of an increased level of cellular turnover (Rao & Venkateswara 2006). At 30 °C, there
was no effect of diet on LDH activities, but at 37 °C the
inclusion of dietary nucleotides exacerbated the effect of heat
on LDH activities, possibly indicating further tissue damage
from fish fed this diet.
Plasma total protein levels were highly consistent at both
temperatures, but at 37 °C a clear reduction in the plasma
total protein of fish fed the NUC diet was observed. This
reduction in plasma total protein levels is consistent with
possible liver damage in fish fed the NUC diet at 37 °C
(Refstie et al. 2006).
No interaction terms among diet or temperature on any of
the plasma enzymology or plasma total protein levels were
observed.

This study demonstrates that there is some potential to
improve the response of barramundi to thermal stress using
dietary strategies. Increasing the protein to energy ratio
provides the best option of those examined to minimize

growth retardation owing to thermal stress. Other alternatives such as dietary HIS and NUC fortification, while
improving growth performance relative to the REF diet at
optimal temperatures, did not provide any benefit under
conditions of heat stress.

The authors acknowledge the support of Peter McCafferty
and Ken Dods at the Chemistry Centre of WA for performing some of the analytical work associated with this
project. They also thank Malcolm McGrath from the Western Australian Animal Health Laboratories for undertaking
the plasma chemistry work. Dr Brian Jones provided editorial input. The financial support of the Australian Centre for
International Agricultural Research is also acknowledged.

Barnes, J.C., Hauler, R.C. & Carter, C.G. (2006) The effect of protein : energy ratio on the growth and protein synthesis of Atlantic
salmon (Salmo salar) at 19 °C. In: Proceedings of the International
Society for Fish Nutrition and Feeding, p. 56, Biarritz, France, 28
May–1 June. INRA, St. Pee Sur Nivelle, France.
Begum, G. (2004) Carbofuran insecticide induced biochemical
alterations in liver and muscle tissues of the fish Clarius batrachus
(linn) and recovery response. Aquat. Toxicol., 66, 83–92.
Boonyaratpalin, M. & Williams, K.C. (2001) Asian seabass. In:
Nutrient Requirements and Feeding of Finfish for Aquaculture
(Webster, C. & Lim, C. eds), pp. 80–90. CABI Publishing, Wallingford, UK.
Breck, O., Bjerkas, E., Campbell, P., Rhodes, J.D., Sanderson, J. &
Waagbo, R. (2005) Histidine nutrition and genotype affect cataract
development in Atlantic salmon, Salmo salar L. J. Fish Dis., 28,
357–371.
Carter, C.G., Bridle, A.R., Ketersky, R.S., Barnes, J.C. & Hauler,
R.C. (2006) Influence of nutrition and thermal stress on protein
synthesis and degradation in fish. In: Proceedings of the International Society for Fish Nutrition and Feeding, p. 57, Biarritz,
France, 28 May–1 June. INRA, St. Pee Sur Nivelle, France.

Folch, J., Lees, M. & Sloane-Stanley, G.H. (1957) A simple method
for the isolation and purification of total lipids from animal
tissues. J. Biol. Chem., 226, 497–509.
Glencross, B.D. (2006) Nutritional management of barramundi,
Lates calcarifer – a review. Aquacult. Nutr., 12, 291–309.
Glencross, B.D. (2008) A factorial growth and feed utilisation model
for barramundi, Lates calcarifer based on Australian production
conditions. Aquacult. Nutr., 14, 360–373.
Helland, S., Grisdale-Helland, B. & Nerland, S. (1996) A simple
method for the measurement of daily feed intake of groups of fish
in tanks. Aquaculture, 139, 156–163.
Katersky, R. & Carter, C.G. (2005) Growth efficiency of juvenile
barramundi, Lates calcarifer, at high temperatures. Aquaculture,
250, 775–780.
Kopp, J., Loos, B., Spilker, G. & Horch, R.E. (2004) Correlation
between serum creatinine kinase levels and extent of muscle
damage in electrical burns. Burns, 30, 680–683.

..............................................................................................

Ó 2009 The Authors
Journal compilation Ó 2009 Blackwell Publishing Ltd Aquaculture Nutrition 16; 343–350


Li, P. & Gatlin, D.M. (2006) Nucleotide nutrition in fish: current
knowledge and future applications. Aquaculture, 251, 141–152.
Maynard, L.A. & Loosli, J.K. (1969) Animal Nutrition, 6th edn.
McGraw-Hill Book Co., New York, NY.
McMeniman, N. (1998) Digestibility of feed ingredients in barramundi. In: Fishmeal Replacement in Aquaculture Feeds for Barramundi. Final Report of Project 93/120-03 (Williams, K.C. ed.),
pp. 16–37. Fisheries Research and Development Corporation,

Canberra, Australia.
Mora, C. & Maya, M.F. (2006) Effect of the rate of temperature
increase of the dynamic method on the heat tolerance of fishes.
J. Thermal Biol., 31, 337–341.
Rao, J. & Venkateswara, A. (2006) Biochemical alterations in
euryhaline fish, Oreochromis mossambicus exposed to sub-lethal
concentrations of an organophosphorus insecticide, monocrotophos. Chemosphere, 65, 1814–1820.
Refstie, S., Glencross, B., Landsverk, T., Sørensen, M., Lilleeng, E.,
Hawkins, W. & Krogdahl, A. (2006) Digestive function and

intestinal integrity in Atlantic salmon (Salmo salar) fed kernel
meals and protein concentrates made from yellow or narrowleafed lupins. Aquaculture, 261, 1382–1395.
Saoudi, M., Abdelmouleh, A., Kammoun, W., Ellouze, F., Jamoussi,
K. & El Feki, A. (2008) Toxicity assessment of the puffer fish
Lagocephalus lagocephalus from the Tunisian coast. C. R. Biol.,
331, 611–616.
Uauy, R., Stringel, G., Thomas, R. & Quan, R. (1990) Effect of
dietary nucleotides on growth and maturation of the developing
gut in the rat. J. Paediatr. Gastroenterol. Nutr., 10, 497–503.
Williams, K.C., Barlow, C.G. & Rodgers, L. (2001) Efficacy of
crystalline and protein-bound amino acids for amino acid enrichment of diets for barramundi/Asian seabass (Lates calcarifer
Bloch). Aquacult. Res., 32, 415–429.
Williams, K.C., Barlow, C.G., Rodgers, L. & Agcopra, C. (2006)
Dietary composition manipulation to enhance the performance of
juvenile barramundi (Lates calcarifer Bloch) reared in cool water.
Aquacult. Res., 37, 914–927.

..............................................................................................

Ó 2009 The Authors

Journal compilation Ó 2009 Blackwell Publishing Ltd Aquaculture Nutrition 16; 343–350


Aquaculture Nutrition
2010 16; 351–358

..........................................................................................

1
1

1

1

doi: 10.1111/j.1365-2095.2009.00671.x

2,3

College of Aquaculture and Fisheries, Can Tho University, Can Tho City, Vietnam; 2 Department of Fisheries – Research
Division, North Beach, WA, Australia; 3 Present address: CSIRO Marine and Atmospheric Research, Cleveland, Qld,
Australia

In this study, we undertook a series of experiments to
assess different faecal-collection methods and also the
digestibilities of a range of feed ingredients when fed
to Tra catfish (Pangasinodon hypothalamus). Apparent
digestibility coefficients for dry matter, energy and nitrogen
for a reference diet were also determined from faeces
collected by way of settlement at 2-h intervals from 2

to 24 h. No significant effect of sample collection time
on the digestibility values was observed. Stripping was
attempted, but was not considered a suitable method for
collecting digesta from Tra catfish. Dissection collection
was also evaluated. Values determined from dissection
collection were lower than those from using settlement.
In a second experiment, a suite of test ingredients was
combined with a reference diet component in a 30 : 70
ratio to determine the ingredients for dry matter, energy
and nitrogen digestibilities. Ingredients included were trashfish, defatted rice bran, wet full-fat rice bran, dried full-fat
rice bran, broken rice, cassava, soybean and fish meals.
Collection of faeces by settlement was used for determining
the digestibility of each of the ingredients. The results
showed that Tra catfish can efficiently digest protein and
energy from a wide range of feed ingredients and nutrient
sources.
KEY WORDS:

catfish, digestibility, faecal collection, methods,
Pangasinodon hypothalamus, Pangasius, rice bran

The culture of Pangasius catfish in the Mekong Delta region is
increasing in terms of both production tonnage and culture
area. The total production of catfish in 2007 was
815 000 tonnes, which accounted for approximately 80% of
the total freshwater aquaculture production of the Mekong
Delta region (MOFI 2008). Of this production, the majority is
of Tra catfish, Pangasius hypothalamus. The average feed cost
typically comprises more than 80% of the total variable production costs in this industry, varying from 73.6% among
farm-made feed category to 92.5% among manufactured

pellet feed farm category (Phuong et al. 2007). However, there
is little data on the nutritional value of most common feed
ingredients used in diets for Tra catfish. The evaluation of the
digestible protein and energy value of feed ingredients is critical to the cost-effective formulation of modern aquaculture
diets and is also an important part of the process in establishing their nutritional value (Glencross et al. 2007). However, it is well known that the faecal-collection method can
influence the digestibility assessment of a diet (Weatherup &
McCracken 1998; Vandenberg & de la Noue 2001; Glencross
et al. 2005). Therefore in this study, we aimed to determine the
feasibility for collection of faecal samples from Pangasinodon
hypothalamus and then used the optimal of the faecal-collection methods to assess the digestible protein and energy value
of a range of feed ingredients widely used in feed formulations
for Tra catfish in the Mekong Delta region of Vietnam.

Received 12 December 2008, accepted 25 February 2009
Correspondence: Dr Brett Glencross, Brett PO Box 20, North Beach 6920.
E-mail:

Experimental fish were hatchery reproduced at Cantho
University. Fish were on-grown to test size ($100 g) in

..............................................................................................

Ó 2009 The Authors
Journal compilation Ó 2009 Blackwell Publishing Ltd


concrete tanks by feeding a commercial pelleted feed
(UniPresident, Dong Nai City, Vietnam) containing 32%
protein. One week prior to start the experiment, fish were
transferred into experimental tank system for acclimation

and continued to fed the same diet prior to the introduction
of the experimental reference diet.

The apparent digestibility coefficients (ADC) were estimated
for a reference (Table 1) and a series of test diets (Cho et al.
1982). The reference diet was formulated to ensure that fish
Table 1 Formulations and composition of the reference diet (all
value are g kg)1)
Ingredients

g kg)1 diet

Fish meal
Soybean meal
Wheat flour
Squid oil
Vitamin and mineral premix1
carboxymethyl cellulose (CMC)
Chromic oxide
Diet proximate specifications
DM contents
Crude protein (g kg)1 DM)
Crude fat (g kg)1 DM)
Ash (g kg)1 DM)
Carbohydrate (g kg)1 DM)
Gross energy (MJ kg)1 DM)

260
200
435

65
10
20
10
954
352
113
74
462
20.9

Vitamin and mineral premix includes (IU kg)1 or g kg)1 of premix): vitamin A, 400 000 IU; vitamin D3, 80 000 IU; vitamin E, 12 g;
vitamin K, 3, 2.4 g; vitamin B1, 1.6 g; vitamin B2, 3.0 g; vitamin B6,
1.0 g; niacin 1.0, vitamin B9, 0.8; vitamin B12, 0.004 g; folic acid,
0.032, biotin, 0.17 g; vitamin C, 60 g; choline, 4.8 g; inositol, 1.5 g;
ethoxyquin, 20.8 g; Copper, 10 g; ferrous iron, 20.0 g; magnesium,
16.6 g; manganese, 2.0 g; zinc, 11 g.
DM, dry matter.

1

Fishmeal
Ingredient specifications
DM (g kg)1)
921
Protein
677
Lipid
65.4
Ash

199
Carbohydrate1
51.9
Energy (MJ kg)1 DM)
18.7

obtained all essential nutrition for the normal growth
according to currently understood Pangasius catfish dietary
requirements (Hien & Yen 2005). The proximate compositions of the reference diet are: 350 g kg)1 protein, 100 g kg)1
fat, 280 g kg)1 starch and 10 g kg)1 chromic oxide. A portion of the reference diet was combined with each test
ingredient in a 70 : 30 ratio. A total of eight test ingredients
were evaluated; these included: defatted rice bran (DRB), wet
full-fat rice bran (WFRB), dried full-fat rice bran (DFRB),
fishmeal, broken rice, cassava meal, defatted soybean meal
and trash-fish (Table 2). The diets were prepared by thoroughly mixing the dry ingredients with oils and then adding
distilled water until a stiff dough was formed. This was then
passed through a screw-press with a 2.5-mm die before being
dried for 12 h at 60 °C and stored at )20 °C until use. Trashfish were included into the diet on a wet basis [allowing for an
estimated 30% inclusion on a dry matter (DM) basis] before
the diets were dried.

For the settlement methodology studies, a series of digestibility tanks of 180 L were used. Each tank had a cylindroconical base sloped at 360 fitted with a 65-mm diameter,
250-mm long collection chamber that tapered into a 12-mm
diameter, 150-mm length of silicone tubing. Continuously
flowing, preheated ($28 °C) freshwater was filtered through
a cotton filter, then a diatomaceous earth filter before
passing on to the experimental tanks at a flow rate of
600 mL min)1. Each tank was aerated using two air–stone
diffusers (Allan et al. 1999). For dissection and stripping
method assessment, fish from the same batch were maintained at the same conditions, but in 600-L square concrete

tanks.

Soybean

TF

WFRB

DFRB

DRB

Cassava

BR

921
457
58.0
76.3
358
18.9

244
731
58.0
164
46.9
18.3

880
134
155
67.1
622
19.1

907
141
208
100
484
18.2

904
165
19.0
105
644
15.3

904
38.0
20.0
40.2
872
15.54

898
84.7

5.8
5.6
903
17.0

Table 2 Composition (g kg)1 DM,
unless otherwise detailed) of key feed
ingredients for Tra catfish

TF, marine trash-fish; WFRB, wet full-fat rice bran; DFRB, dried full-fat rice bran; DRB, defatted
rice bran; BR, broken rice.
Fishmeal: Kiengiang Fish meal Company, Kiengiang province, Vietnam; Soybean: RAJA Fat and
feeds Private limited, India; Trash-fish: Local fishermen, Kiengiang province, Vietnam; WFRB,
DFRB, BR: Broken rice, cassava meal: Gentraco Feed, Cantho province, Vietnam; DRB: Cai Lan Oils
& Fats Industries Company Ltd, Can Tho Branch, Cantho City, Vietnam.
1
Carbohydrate based on dry matter (DM) ) (protein + ash + lipid).

..............................................................................................

Ó 2009 The Authors
Journal compilation Ó 2009 Blackwell Publishing Ltd Aquaculture Nutrition 16; 351–358


This experiment was conducted in three settlement tanks.
Mass of faeces collected, ADC of DM, energy and nitrogen
for the reference diet were calculated and compared with
faeces collected by every 2 h after feeding (2, 4, 6, 8, 10, 12,
14, 16, 18, 20, 22, 24 h). Fish were stocked 7 days prior to the
start of the faecal-collection period to allow for acclimatization to experimental conditions. Twenty fish (84 g fish)1)

were stocked into each tank. During this period, fish were fed
the reference diet (Table 1). Fish were fed to excess using by
hand once daily at 8 AM. Within 1 h after feeding, all uneaten
food was removed, and the walls of the tank and the settlement chamber were thoroughly cleaned to remove any faeces,
uneaten food or bacterial slime. The silicone tubing into
which the faeces settled was packed in ice and kept at )4 °C
prior to the removal of faeces to reduce bacterial proliferation, which can affect the composition of faeces (Spyridakis
et al. 1989). During the experiment, mean water temperature
was 26.8–28.2 °C, dissolved oxygen was always maintained
above 6.3 mg L)1 and water pH ranged from 7.7 to 8.1.

The ADC of DM, energy and nitrogen for the reference diet
and two test ingredients (fish meal and soybean meal) were
calculated and compared using either settlement, stripping or
dissection faecal-collection technique. Fish were stocked
7 days prior to the start of the faecal-collection period to
allow for acclimation to the experimental conditions and
each of the diets. Twenty fish ($91 g) were allocated to each
of nine settlement tanks. Eighteen large 600-L square concrete tanks were stocked with 50 fish of the same sample size
as used in the nine settlement tanks.
During this acclimation period, fish were fed each diet to
excess using by hand once daily at 8 AM. On the eighth day,
after the 12th hour of fish feeding, faecal collection was
commenced. For the settlement technique, faecal samples
from different days but within each tank were pooled for
analysis. For dissection techniques, all 50 fish within each
tank were killed, the distal intestine dissected and faeces were
expelled with gentle pressure into sample container. For
stripping, all 50 fish were netted from each tank, the fish were
anaesthetized using MS-222, and then gentle pressure applied

to the abdomen near the distal intestine of the fish to expel
faeces onto the fingertips. The faeces were placed into a
sample container and stored at )20 °C before being prepared

for analysis. All fish were stripped only once. During
experiments, dissolved oxygen was always above 5 mg L)1,
pH between 7.7 and 8.3, and water temperature ranged from
29 to 30 °C.

This study consisted of nine treatments each allocated to
three randomly selected, replicate tanks. A sample of the
reference diet was combined with each test ingredient in a
70 : 30 ratio (Table 3). A total of eight test ingredients and
one reference diet were evaluated. The test ingredients
included; DRB, WFRB, DFRB, fishmeal, broken rice, cassava meal, soybean meal and trash-fish (Table 2). Faeces
were collected by settlement over a 12-h period. The ADC of
each of the test ingredients was calculated based on defining
the ADC of test diets and the ADC of reference diet based on
the proportion of reference diet and test ingredient in each
test diet, and the proportion of nutrient/energy contributed
by each test ingredient in test diet (Sugiura et al. 1998).
Twenty fish ($96 g fish)1) were stocked into each tank (total
27 tanks). After 7 days acclimatization, faeces were collected
and pooled for each tank for 7 days. Faeces were stored at
)20 °C before being dried for analysis. Mean water temperature during the experiment was 30.3 °C, dissolved oxygen always above 5.5 mg L)1 and pH between 7.5 and 8.0.
Table 3 Formulations of the experimental diets (all values are in
g kg)1)
Diet
Ingredients

1

2

3

4

5

6

7

8

9

Fish meal
260 182 182 182 182 182 182 182 182
Soybean meal
480 336 336 336 336 336 336 336 336
Wheat flour
200 140 140 140 140 140 140 140 140
Squid oil
20 14 14 14 14 14 14 14 14
Chromic oxide
10
7
7

7
7
7
7
7
7
CMC
20 14 14 14 14 14 14 14 14
Vit-premix1
10
7
7
7
7
7
7
7
7
Fish meal
– 300
–
–
–
–
–
–
–
Soy meal
–
– 300

–
–
–
–
–
–
Trash-fish
–
–
– 300
–
–
–
–
–
Wet fat rice bran
–
–
–
– 300
–
–
–
–
Dry full-fat rice bran
–
–
–
–
– 300

–
–
–
Defatted rice bran
–
–
–
–
–
– 300
–
–
Cassava
–
–
–
–
–
–
– 300
–
broken rice
–
–
–
–
–
–
–
– 300

Vitamin and mineral premix includes (IU kg)1 or g kg)1 of premix): vitamin A, 400 000 IU; vitamin D3, 80 000 MIU; vitamin E,
12 g; vitamin K, 3, 2.4 g; vitamin B1, 1.6 g; vitamin B2, 3.0 g; vitamin B6, 1.0 g; niacin 1.0, vitamin B9, 0.8; vitamin B12, 0.004 g; folic
acid, 0.032, biotin, 0.17 g; vitamin C, 60 g; choline, 4.8 g; inositol,
1.5 g; ethoxyquin, 20.8 g; Copper, 10 g; ferrous iron, 20.0 g; magnesium, 16.6 g; manganese, 2.0 g; zinc, 11 g.

1

..............................................................................................

Ó 2009 The Authors
Journal compilation Ó 2009 Blackwell Publishing Ltd Aquaculture Nutrition 16; 351–358


Diet and faecal samples were analysed for DM, chromium,
ash, phosphorus, nitrogen and gross energy content. DM was
calculated by gravimetric analysis following oven drying at
105 °C for 24 h. Energy content was determined using ballistic bomb calorimetry. Protein levels were calculated from
the determination of total nitrogen by kjeldhal analysis,
based on n · 6.25. Crude fat levels were assessed using the
acid-hydrolysis Soxhlet method. Chromic oxide levels were
determined by spectrophotometric analysis following heating
in kjeldhal system with nitric acid and perchloric acid. Carbohydrate was determined as the difference in DM content
minus protein, ash and fat. All of these determinations were
conducted according to the methods specified by the AOAC
(2000).
Differences in the ratios of the parameters of DM, protein
or gross energy relative to chromium, in the feed and faeces
in each treatment were calculated to determine the apparent
digestibility coefficient (ADCdiet) for each of the nutritional
parameters examined in each diet based on the following

formula.
Diet 1 in each experiment is the referent diet. The apparent
digestibility coefficient (ADCdiet) of each specific nutritional
variable is based on the following formula (Eqn 1):
ADCdiet ¼ 1 À



Crdiet  Nutrientfaeces
Crfaeces  Nutrientdiet

ð1Þ

In this equation, the terms Crdiet and Crfaeces represent the
chromium content of the diet and faeces respectively, and
Nutrientdiet and Nutrientfaeces represent the nutritional
parameter of concern (e.g. protein or energy) in the diet and
faeces respectively. With this formula, values range from 0 to
1. To achieve a percent apparent digestibility, the values
should be multiplied by 100.
The Nutr.ADingredient is the digestibility of a given nutrient
from the test ingredient included in the test diet at 30%.
ADNtest is the apparent digestibility of the nutrient of
interest in the test diet. ADC reference is the apparent
digestibility of the same nutrient from the reference diet,
which makes up 70% of the test diet (Eqn 2):

ADref is the apparent digestibility of the basal diet, which
makes up 70% of the test diet. NutrIngredient, Nutrtest and
Nutrref are the level of the nutrient of interest in the ingredient, test diet and basal diet respectively (Sugiura et al.

1998). Where the inclusion level of the test ingredient varies,
then the new ratio needs to be included in this equation in
place of the 70% and 30% values.

The data were subjected to one-way analysis of variance
(ANOVA; P < 0.05). Significant differences between treatment
means were compared using DuncanÕs multiple range test at
the 5% level of significance. Data normality and homogeneity were checked using SPSS (SPSS Inc., Chicago, Illinois,
USA) tests, version 13.0.

Faecal quantities based on the amounts of faeces collected by
way of settlement increased from the 2- to 14-h time point
after feeding. From the collections after 12- to 18-h time
point, the faecal quantities collected were higher compared
with the other time point, e.g. 2- to 14-h time point (Fig. 1).
After 18-h time point, the faecal quantities collected diminished. The ADC for DM calculated using faeces collected for
each separate 2-h interval showed no significant differences
across time.

1.00
0.90
0.80
0.70
0.60

ADC

0.50

Amount (g/tank)

0.40
0.30
0.20
0.10

ðADtest ÂNutrtest À ðADref  Nutrref Â70%ÞÞ
À
Á
Nutr.ADingredient ¼
30% Â NutrIngredient

0.00
0

5

10
15
Time post feeding (h)

20

25

ð2Þ
In Eqn 2, the Nutr.ADingredient is the digestibility of a given
nutrient from the test ingredient included in the test diet at
30%. ADtest is the apparent digestibility of the test diet.

Figure 1 Quantities of faeces collected and dry matter digestibility
over varying settlement periods postfeeding. No significant differences in dry matter digestibility (apparent digestibility coefficient,
ADC) were observed over time. Each data point represents n = 3.

..............................................................................................

Ó 2009 The Authors
Journal compilation Ó 2009 Blackwell Publishing Ltd Aquaculture Nutrition 16; 351–358


Stripping was identified as not a suitable method for collecting digesta from Tra catfish. There was difficulty in
obtaining sufficient sample because of the structure of the fish
intestinal tract and abdominal muscle layer. However, sufficient samples and faeces/digesta was able to be collected
using either the settlement or dissection methods. Significant
differences between the settlement and dissection-collection
methods were noted on the digestibility traits of the diets.
DM, protein and energy digestibilities of the diets calculated
from samples collected using dissection were consistently
lower than those obtained using settlement collection.
However, the digestibility coefficients of the test ingredients
(fish meal and soybean meal) were unaffected by faecalcollection method (Table 4).

Ingredient composition The ingredients evaluated in this
study had a range of compositions (Table 2). Among
the protein source ingredients, the fish meal and trash-fish
had significantly different levels of protein (677 and
731 g kg)1 DM respectively) and both were significantly
higher than that of the soy protein (457 g kg)1 DM).
The DRB had a higher protein level (165 g kg)1 DM),
than wet rice bran and dry rice bran protein levels (134

and 141 g kg)1 DM). The protein level of cassava meal

Table 4 Comparison of faecal-collection methods on diet and
ingredient digestibility parameters
Collection
methods

Dry matter
(%)

Reference diet digestibility
Dissection
80.2 ± 0.6a
Settlement
89.2 ± 0.9b
Fish meal diet digestibility
Dissection
81.8 ± 0.7a
Settlement
88.7 ± 0.6b
Soybean diet digestibility
Dissection
78.4 ± 3.3a
Settlement
87.0 ± 1.1b
Fish meal ingredient digestibility
Dissection
85.5 ± 2.5a
Settlement
87.7 ± 1.8a

Soybean ingredient digestibility
Dissection
74.2 ± 11a
Settlement
81.9 ± 3.7a

Protein
(%)

Energy
(%)

84.6 ± 0.7a
94.8 ± 1.1b

86.7 ± 0.4a
92.8 ± 0.6b

88.9 ± 1.1a
95.5 ± 0.8b

88.8 ± 0.5a
93.1 ± 0.3b

86.7 ± 2.8a
94.6 ± 1.4b

85.5 ± 2.2a
91.3 ± 0.7b

82.2 ± 2.4a
96.12 ± 1.6b

93.6 ± 1.5a
93.9 ± 1.1a

72.8 ± 7.7a
94.4 ± 3.9b

82.7 ± 7.2a
88.0 ± 2.4a

Values are means ± SD (n = 3). Different superscripts within rows
indicate significant differences between two collection methods at
P < 0.05.

was the lowest (38 g kg)1 DM). In addition, the DFRB
also had higher level of lipid (208 g kg)1 DM). The rice
bran group had carbohydrate levels >484 g kg)1 DM,
which was less than that of the broken rice and cassava
meal (872–903 g kg)1 DM).
Ingredient assessment based on settlement collection Based
on faecal samples collected using settlement techniques,
DM digestibilities of the ingredients varied substantially
(Table 5). For protein feed ingredients, the total levels of
digestible DM was the lowest for SBM (81.9%) and the
highest for both FM and trash-fish (87.7%), which was
found to be a significant difference (P < 0.05) among those
ingredients. DM digestibilities were significantly higher for
broken rice and cassava meal (83.2–90.7%) than rice bran,

which ranged between 57% and 82%. In rice bran group,
WRB had a significantly (P < 0.05) higher DM digestibility
(82%) in all. Digestibility coefficients for energy were significantly higher for fish meal and trash-fish than for SBM.
Protein digestibility was high for all protein ingredients
94.4–96.1% and not significantly different (P > 0.05)
among those ingredients. However, protein digestibility of
the Cassava meal was the poorest (35.8%) of all the
ingredients evaluated. Protein digestibility was similar for
the remaining ingredients (65.1–70.4%) and were not significantly different (P > 0.05).

This study examined some of the methodological considerations for undertaking digestibilities studies with Pangasius
catfish. The practicalities and results from different faecalcollection methods, the time of postfeeding for optimal collection and the nutritive value of some key feed ingredients
were all examined.

Of the three methods for faecal/digesta collection attempted,
stripping was considered an ineffective method of sampling
digesta from catfish. This was because of difficulty in
obtaining sufficient sample due to the structure of the fish
intestinal tract and abdomen muscle layer. Attempts to
obtain digesta by stripping were also abandoned by Allan
et al. (1999), who were also unable to collect sufficient
quantities from silver perch (Bidyanus bidyanus). Despite
these problems, it is acknowledged that stripping and dissection are the preferred methods for collecting faecal/digesta
samples for fish species, especially where the faeces are

..............................................................................................

Ó 2009 The Authors
Journal compilation Ó 2009 Blackwell Publishing Ltd Aquaculture Nutrition 16; 351–358

Digestibility
parameter
Dry matter
Protein
Energy

Fishmeal
de

87.7
96.1c
93.9d

Soybean
c

81.9
94.4c
88.0c

TF

WFRB
e

89.1
94.5c
96.4d

c

81.5
67.6b
84.2c

DFRB
b

62.9
70.4b
65.6b

DRB
a

56.7
66.4b
54.5a

Cassava
cd

83.2
35.8a
84.4c

BR
90.7e
65.1b

87.1c

Table 5 Apparent digestibility (%)
specifications of test ingredients as
determined using settlement faecalcollection methods

Values are means (n = 3). Different superscripts within rows indicate significant differences
between means among ingredients at P < 0.05.
TF, marine trash-fish; WFRB, wet full-fat rice bran; DFRB, dried full-fat rice bran; DRB, defatted
rice bran; BR, broken rice.

loosely bound or contained substantial levels of carbohydrates (McMeniman & Sands 1996; Glencross et al. 2005).
The ADCs calculated using digesta obtained by dissection
were considerably lower than those calculated using faeces
collected using the settlement method. This was consistent
with other reports comparing active and passive collection
methods (Vandenberg & de la Noue 2001; Glencross et al.
2005). Settlement collection was used as the preferred method
primarily due to the ability to collect a larger faecal sample
size without interfering or sacrificing fish and in the minimum
time. However, collecting faeces after they have been voided
can result in leaching of DM and nutrients and potentially
lead to overestimation of digestibility values. Collection
facilities that ensure rapid settlement of faeces (Cho &
Kaushik 1990; Satoh et al. 1992; Hajen et al. 1993) or continuous filtration from the water column (Choubert et al.
1982; Spyridakis et al. 1989) have been used as an attempt to
reduce this problem.
Significant effects of faecal-collection method were also
noted on the digestibility assessment of the test ingredients
(fish meal and soybean meal) in this study. This is consistent

with the findings of Glencross et al. (2005) who also showed
that faecal-collection method affected ADCs of both the diets
and ingredients fed to rainbow trout.

The similarity between the ADCs calculated from faeces
collected at 2-h intervals, and those calculated using faeces
collected over a 24-h period cumulatively after feeding,
supports that leaching from faeces in the collecting chamber
was not a significant problem over a 24-h period and that
there was no advantage in collecting samples more frequently
than a daily sample. Similarly, Satoh et al. (1992) found
minimal differences in lipid or protein digestibility when
faeces from rainbow trout were collected using settlement 3,
6, 9, 12 or 15 h after feeding. Similar results were also
obtained when faeces from silver perch were collected using
settlement periods of 2–18 h after tanks were cleaned (Allan
et al. 1999).

The composition of the feedstuffs evaluated in this study is
consistent with the values reported in earlier publications
(NRC 1993; Usmani et al. 2003) for similar raw materials.
The rice brans, cassava and broken rice are clearly, based on
their low protein levels, primarily intended as potential energy
sources, whereas the trash-fish, fish meal and soybean meal
constitute the main protein sources. The use of digestibility
information in least-cost diet formulation assumes that the
ADCs for separate ingredients are additive. This was confirmed for rainbow trout by separately measuring digestibility
of component ingredients of a reference diet and then comparing the sum of these individual components on a proportional basis with direct measurement of the complete diet
(Cho et al. 1982). Results show that catfish have the tendency
to digest DM and protein in feedstuffs of animal origin more

efficiently than DM in feedstuffs of plant origin. This is suggestive of a limited ability to digest non-starch polysaccharides by Pangasius catfish. The DM ADC for fish meal was
87.7% while that for soybean meal was only 81.9%. Sullivan
& Reigh (1995) also observed that the DM digestibility of
Menhaden fish meal in hybrid striped bass (Morone soxetilis · Morone chrysops) was higher than that of soybean meal.
However, the present study also found high digestibility of
protein from soybean meal as well as fishmeal and trash-fish
sources. Similar observations have also been reported for
channel catfish, Ictalurus punctatus (Wilson & Poe 1985).
Lorico-Querijero & Chiu (1989) reported high digestibility of
both plant and animal protein sources for tilapia, Oreochrimis
niloticus, whereas in other studies on Mystus nemurus (Khan
1994) and Cyprius carpio (Degani et al. 1997), digestibility
values for fishmeal were found to be higher than that of
soybean meal. Allan et al. (1999) reported that silver perch
(B. bidyanus) has higher DM ADCs for fishmeal higher than
that of soybean, but protein ADC values were high and no
significant differences were observed among those ingredients.
Protein digestibility observed for rice bran in the present
study was similar to values reported for other species Clarias
batrachus and Clarias gariepinus (Usmani et al. 2003),

..............................................................................................

Ó 2009 The Authors
Journal compilation Ó 2009 Blackwell Publishing Ltd Aquaculture Nutrition 16; 351–358


Channel catfish (Wilson & Poe 1985). The relatively low
values for protein digestibility are suggested to be artefacts
from the high levels of indigestible fibre present in the rice

bran acting as bulking agents with these ingredients and are
consistent with the low DM and energy digestibilities also
seen for these ingredients. The low protein digestibility of the
cassava is unusual and suggests the influence of either antinutritional factors such as protease inhibitors (Francis et al.
2001) or protein damage from cooking (Glencross et al.
2004). The high DM and energy digestibilities of broken rice,
cassava and wet rice bran that were noted in the present study
support that these ingredients are potentially useful energy
sources for use in Pangasisus diets. In contrast, the digestibility of DM and energy from DFRB and DRB were low and
indicative that these ingredients are less suitable as energy
sources.

In conclusion, the collection of faeces by settlement was the
preferred method for determining digestibility in Tra
(Pangasius) catfish. The results show that Tra catfish can
digest protein and energy from a wide range of feed ingredients and nutrient sources. These data can be used not only
to better define nutrient and energy requirements but also to
better formulate nutritionally efficient, cost-effective feeds for
this species. However, further work is required to identify
and evaluate the nutritional value of additional protein
sources other than trash-fish and soybean and fish meals.

Allan, G.L., Rowland, S.J., Parkinson, S., Stone, D.A.J. & Jantrarotai, W. (1999) Nutrient digestibility for juvenile silver perch
Bidyanus bidyanus: development of methods. Aquaculture, 170,
131–145.
AOAC (2000) Official Methods of Analysis of the Association of
Official Analytical Chemists. AOAC, Washington, DC.
Cho, C.Y. & Kaushik, S.J. (1990) Nutritional energetics in fish:
energy and protein utilisation in rainbow trout (Salmo gairdnerii).
World Rev. Nutr. Diet., 61, 132–172.

Cho, C.Y., Slinger, S.J. & Bayley, H.S. (1982) Bioenergetics of salmonid fishes: energy intake, expenditure and productivity. Comp.
Biochem. Physiol., 73B, 25–41.
Choubert, G., De la Noue, J. & Luquet, P. (1982) Digestibility in
fish: improved device for the automatic collection of feces. Aquaculture, 29, 185–189.
Degani, G., Viola, S. & Yehuda, Y. (1997) Apparent digestibility
coefficient of protein sources for carp, Cyprinus carpio L. Aquacult.
Res., 28, 23–28.
Francis, G., Makkar, H.P.S. & Becker, K. (2001) Antinutritional
factors present in plant-derived alternate fish feed ingredients and
their effect in fish. Aquaculture, 199, 197–227.

Glencross, B.D., Hawkins, W.E. & Curnow, J.G. (2004) Nutritional
assessment of Australian canola meals. I. Evaluation of canola oil
extraction method, enzyme supplementation and meal processing
on the digestible value of canola meals fed to the red seabream
(Pagrus auratus, Paulin). Aquacult. Res., 35, 15–24.
Glencross, B.D., Hawkins, W.E., Evans, D., McCafferty, P., Dods,
K., Maas, R. & Sipsas, S. (2005) Evaluation of the digestible
value of lupin and soybean protein concentrates and isolates
when fed to rainbow trout, Oncorhynchus mykiss, using either
stripping or settlement faecal collection methods. Aquaculture,
245, 211–220.
Glencross, B.D., Booth, M. & Allan, G.L. (2007) A feed is only as
good as its ingredients – a review of ingredient evaluation for
aquaculture feeds. Aquacult. Nutr., 13, 17–34.
Hajen, W.E., Beames, R.M., Higgs, D.A. & Dosanjh, B.S. (1993)
Digestibility of various feedstuffs by post-juvenile chinook salmon
Oncorhynchus tshawytscha.in sea water: I. Validation of technique.
Aquaculture, 112, 321–332.
Hien, T.T.T. & Yen, D.T. (2005) Study on Nutrition Requirement for 3

Species Cast Fish Pangasinodon bocourti, P. conchophilusand and
P. hypophthalmus. Ministry of Education and Academic Project, 60
pp (in Vietnamese). CanTho University, CanTho City, Vietnam.
Khan, M.S. (1994) Apparent digestibility coefficients for common
feed ingredients in formulated diets for tropical catfish, Mystus
nemurus (Cuvier & Valenciennes). Aquac. Fish. Manage., 25, 167–
174.
Lorico-Querijero, B.V. & Chiu, Y.N. (1989) Protein digestibility
study in Oreochromis niloticus using chromic oxide indicator.
Asian Fish. Sci., 2, 177–191.
McMeniman, N.P. & Sands, N. (1996) Estimation of in vivo
digestibility of diets to barramundi Lates calcarifer. In: Second
World Fisheries Congress Proceedings, Vol. 1 (Hancock, D.A. &
Beumer, J.P. eds), pp. 100–101. CSIRO Publishing, Melbourne,
Vic., Australia.
MOFI (Ministry of Fisheries) (2008) Annual Report by the Ministry
of Fisheries (MOFI). MOFI, Hanoi, Vietnam. (In Vietnamese).
NRC (National Research Council) (1993) Nutrient Requirements of
Fish. National Academy Press, Washington, DC, 114 pp.
Phuong, N.T., Sinh, L.X., Thinh, N.Q., Chau, H.H., Anh, C.T. &
Hau, N.M. (2007) Economics of aquaculture feeding practices:
Viet Nam. In: Economics of Aquaculture Feeding Practices in
Selected Asian Countries (Hasan, M.R. Ed.), pp. 183–205. Technical Paper No.505, FAO, Rome.
Satoh, S., Cho, C.Y. & Watanabe, T. (1992) Effect of faecal retrieval
timing on digestibility of nutrients in rainbow trout diet with the
Guelph and TUF feces collection systems. Nippon Suisan Gakkai
Shi, 58, 1123–1127.
Spyridakis, R., Metailler, J., Gabaudan, J. & Riaza, A. (1989)
Studies on nutrient digestibility in European seabass Dicentrarchus
labrax: I. Methodological aspects concerning digesta collection.

Aquaculture, 77, 61–70.
Sugiura, S.H., Dong, F.M., Rathbone, C.K. & Hardy, R.W. (1998)
Apparent protein digestibility and mineral availabilities in various feed ingredients for salmonid feeds. Aquaculture, 159, 177–
202.
Sullivan, J.A. & Reigh, R.C. (1995) Apparent digestibility of selected
feedstuffs in diets for hybrid striped bass (Morone saxatilis female
x Morone chrysops male). Aquaculture, 138, 313–322.
Usmani, N., Jafri, A.K. & Khan, M.A. (2003) Nutrient digestibility
studies in Heteropneustes fossilis (Bloch), Clarias batrachus
(Linnaeus) and C. gariepinus (Burchell). Aquacult. Res., 34, 1247–
1253.

..............................................................................................

Ó 2009 The Authors
Journal compilation Ó 2009 Blackwell Publishing Ltd Aquaculture Nutrition 16; 351–358


Vandenberg, G.W. & de la Noue, J. (2001) Apparent digestibility
comparison in rainbow trout (Oncorhynchus mykiss) assessed using
three methods of faeces collection and three digestibility markers.
Aquacult. Nutr., 7, 237–245.
Weatherup, R.N. & McCracken, K.J. (1998) Comparison of
the estimation of digestibility of two diets for rainbow

trout, Oncorhynchus mykiss (Walbaum), using two markers
and two methods of faeces collection. Aquacult. Res., 29, 527–
533.
Wilson, R.P. & Poe, W.E. (1985) Apparent digestible protein and
energy coefficients of common feed ingredients for channel catfish.

Prog. Fish-Cult., 47, 154.

..............................................................................................

Ó 2009 The Authors
Journal compilation Ó 2009 Blackwell Publishing Ltd Aquaculture Nutrition 16; 351–358


Aquaculture Nutrition
doi: 10.1111/j.1365-2095.2009.00673.x

2010 16; 359–369

..........................................................................................

1

2

3
1

1
1

1

3

1

Laborato´rio de Enzimologia (LABENZ), Departamento de Bioquı´mica; Laborato´rio de Imunopatologia Keizo Asami (LIKA),
Universidade Federal de Pernambuco, Cidade Universita´ria, Recife – PE, Brazil; 2 Embrapa Meio-Norte, Parnaı´ba – PI,
Brazil; 3 Departamento de Bioquı´mica, Universidade Federal de Sa˜o Paulo, Sao Paulo – SP, Brazil

The aim of this study was to examine proteinases and peptidases from the hepatopancreas of two sub-adult stages of
Farfantepenaeus subtilis: SAS6 (5.93 ± 0.69 g wet weight)
and SAS13 (13.26 ± 0.60 g wet weight). Trypsin and chymotrypsin activity was higher in the extract from the SAS6
individuals (P < 0.05). The highest activity among aminoacyl-b-naphthylamide substrates was found using alanine-,
arginine-, leucine- and lysine-b-naphthylamide. There was a
positive correlation between the recommended concentration
of essential amino acids in penaeid shrimp feed and aminopeptidase activity in both sub-adult stages. Proteolytic
activity of F. subtilis was strongly inhibited by specific trypsin
inhibitors. The optimal temperature for trypsin, chymotrypsin and leucine aminopeptidase activity was between 45
and 55 °C. Six and seven bands were found in caseinolytic
zymograms for SAS6 and SAS13 respectively. All bands were
inhibited by phenylmethylsulfonyl fluoride in both sub-adult
stages. The use of tosyl-lysine-chloromethyl-ketone and
benzamidine caused strong inhibition of the proteolytic
bands. Trypsin and chymotrypsin activity was the main difference observed between the protease pattern of SAS6 and
SAS13 F. subtilis.
KEY
WORDS: aminopeptidases,
chymotrypsin, digestive
enzymes, Farfantepenaeus subtilis, southern brown shrimp,
trypsin

Received 6 November 2008, accepted 20 March 2009
Correspondence: Ranilson S. Bezerra, Laborato´rio de Enzimologia –
LABENZ, Departamento de Bioquı´mica, Universidade Federal de Pernambuco, Cidade Universita´ria, Recife – PE 50670-420, Brazil. E-mail:

..............................................................................................

Ó 2009 Blackwell Publishing Ltd
No claim to original US government works

The southern brown shrimp, Farfantepenaeus subtilis, is native
to the Atlantic coast of Central and South America, from
Cuba down to Rio de Janeiro, and was one of the first species
to be farmed in Brazil, along with Farfantepenaeus brasiliensis,
Farfantepenaeus paulensis and Litopenaeus schmitti.
The southern brown shrimp exhibits benthic omnivorous
opportunistic feeding habits under semi-intensive conditions,
although polychaetes and calanoid copepods seem to be
favoured during all growth stages (Nunes & Parsons 2000).
Despite its farming potential and attractive market features, the culture of F. subtilis in semi-intensive conditions in
Brazil has failed mainly due to low yields. Studies carried out
by Brazilian farmers report a food conversion ratio ranging
from 2.88 to 3.44 and a poor growth performance, thus
generating low productivity. The growth rate slows after the
shrimp reach 6 g of body weight. This suggests that the poor
results may be related to nutritional problems and ontogenetic changes in the digestive enzyme metabolism (Maia &
Nunes 2003).
Comprehension of digestion physiology and nutrient
digestibility remains a problem for the culture of F. subtilis.
Knowledge concerning the digestive system of this species
can provide information applicable to food utilization. Thus,
the identification and characterization of digestive enzymes
during shrimp growth is an important step towards understanding the digestive mechanisms and formulating feeds that

promote better growth responses, as feed can be designed
according to the digestive capacity (Lo´pez-Lo´pez et al. 2005).
A number of studies have indicated properties of digestive enzymes in shrimp and other crustaceans, such as
proteases, carbohydrases, lipases and the digestibility of
feed ingredients (Lemos et al. 2000, 2004; Co´rdova-Murueta